Estimating a movement of a hybrid-behavior vehicle
A method for predicting a movement of a hybrid behavior vehicle, the method may include receiving video streams of driving scenes; analyzing the video streams by creating an environment-based behavioral model of each hybrid-behavior vehicle of the vehicles to provide multiple environment-based behavioral models; wherein each environment-based behavioral model associates environmental elements located at an environment of a hybrid-behavior vehicle to one or more predicted behaviors of the hybrid behavior vehicle; and responding to the creation of the multiple environment-based behavioral models.
Latest AUTOBRAINS TECHNOLOGIES LTD Patents:
This application claims priority from 61/827,112 filing date Mar. 31, 2019 and from U.S. provisional patent Ser. No. 62/825251 filing date Mar. 28, 2019 which is incorporated herein by reference.
BACKGROUNDA hybrid-behavior vehicle is a vehicle that is controlled by its user to exhibits, at some point of time a movement of a vehicle and also to exhibit, at other points in time a movement of a pedestrian.
Non-limiting examples of hybrid behavior vehicles include unmotorized, bicycles, motorized bicycles, motorized skateboards, unicycles, hoverboards, kick scooters, and the like.
A hybrid-behavior vehicle is controlled by its user to move in a manner that defies traffic rules. For example—the hybrid-behavior vehicle may move along the road and then may move along the sidewalk. Users of the hybrid-behavior vehicle tend to ignore traffic rules, traffic signs, traffic lights, and may move their hybrid-behavior vehicle regardless of lanes and/or lane direction.
The movement of a hybrid-behavior vehicle is usually characterized by rapid changes of direction, path of propagation, and is hard to predict.
The movement of a hybrid-behavior vehicle is also dependent upon the parameters of its components (chassis, motor—if such exists, brakes, and the like), and the rapid introduction of new and unknown hybrid behavior vehicles to the market also complicates the prediction of the movements of the hybrid behavior vehicle.
There is a growing to predict the movement of hybrid behavior vehicles.
SUMMARYThere may be provided a method for predicting a movement of a hybrid behavior vehicle, the method may include receiving video streams of driving scenes; analyzing the video streams by creating an environment-based behavioral model of each hybrid-behavior vehicle of the vehicles to provide multiple environment-based behavioral models; wherein each environment-based behavioral model associates environmental elements located at an environment of a hybrid-behavior vehicle to one or more predicted behaviors of the hybrid behavior vehicle; and responding to the creation of the multiple environment-based behavioral models.
There may be provided a non-transitory computer readable medium for predicting a movement of a hybrid behavior vehicle, the non-transitory computer readable medium stores instructions for receiving video streams of driving scenes; analyzing the video streams by creating an environment-based behavioral model of each hybrid-behavior vehicle of the vehicles to provide multiple environment-based behavioral models; wherein each environment-based behavioral model associates environmental elements located at an environment of a hybrid-behavior vehicle to one or more predicted behaviors of the hybrid behavior vehicle; and responding to the creation of the multiple environment-based behavioral models.
There may be provided a method for driving an autonomous vehicle, the method may include sensing an environment of the autonomous vehicle by one or more autonomous vehicle sensor; detecting, in the environment of the autonomous vehicle, environmental elements and a hybrid-behavior vehicle; predicting one or more predicted behaviors of the hybrid behavior vehicle based on the environmental elements and an environment-based behavioral model of the hybrid-behavior vehicle; and driving the autonomous vehicle based, at least in part, on the one or more predicted behaviors of the hybrid behavior vehicle.
There may be provided a method for assisting a driver of a vehicle, the method may include sensing an environment of the vehicle by one or more vehicle sensor; detecting, in the environment of the vehicle, environmental elements and a hybrid-behavior vehicle; estimating an estimated future movement of the hybrid-behavior vehicle based on the environmental elements and an environment-based behavioral model of the hybrid-behavior vehicle; and performing at least one out of providing to the driver an indication about the one or more predicted behaviors of the hybrid behavior vehicle to the driver; and calculating a suggested propagation path of the vehicle, based on the one or more predicted behaviors of the hybrid behavior vehicle, and providing to the driver an indication about the suggested propagation path of the vehicle.
There may be provided a non-transitory computer readable medium for driving an autonomous vehicle, the non-transitory computer readable medium that stores instructions for sensing an environment of the autonomous vehicle by one or more autonomous vehicle sensor; detecting, in the environment of the autonomous vehicle, environmental elements and a hybrid-behavior vehicle; predicting one or more predicted behaviors of the hybrid behavior vehicle based on the environmental elements and an environment-based behavioral model of the hybrid-behavior vehicle; and driving the autonomous vehicle based, at least in part, on the one or more predicted behaviors of the hybrid behavior vehicle.
There may be provided a non-transitory computer readable medium for assisting a driver of a vehicle, the non-transitory computer readable medium that stores instructions for sensing an environment of the vehicle by one or more vehicle sensor; detecting, in the environment of the vehicle, environmental elements and a hybrid-behavior vehicle; estimating an estimated future movement of the hybrid-behavior vehicle based on the environmental elements and an environment-based behavioral model of the hybrid-behavior vehicle; and performing at least one out of providing to the driver an indication about the one or more predicted behaviors of the hybrid behavior vehicle to the driver; and calculating a suggested propagation path of the vehicle, based on the one or more predicted behaviors of the hybrid behavior vehicle, and providing to the driver an indication about the suggested propagation path of the vehicle.
The embodiments of the disclosure will be understood and appreciated more fully from the following detailed description, taken in conjunction with the drawings in which:
In the following detailed description, numerous specific details are set forth in order to provide a thorough understanding of the invention. However, it will be understood by those skilled in the art that the present invention may be practiced without these specific details. In other instances, well-known methods, procedures, and components have not been described in detail so as not to obscure the present invention.
The subject matter regarded as the invention is particularly pointed out and distinctly claimed in the concluding portion of the specification. The invention, however, both as to organization and method of operation, together with objects, features, and advantages thereof, may best be understood by reference to the following detailed description when read with the accompanying drawings.
It will be appreciated that for simplicity and clarity of illustration, elements shown in the figures have not necessarily been drawn to scale. For example, the dimensions of some of the elements may be exaggerated relative to other elements for clarity. Further, where considered appropriate, reference numerals may be repeated among the figures to indicate corresponding or analogous elements.
Because the illustrated embodiments of the present invention may for the most part, be implemented using electronic components and circuits known to those skilled in the art, details will not be explained in any greater extent than that considered necessary as illustrated above, for the understanding and appreciation of the underlying concepts of the present invention and in order not to obfuscate or distract from the teachings of the present invention.
Any reference in the specification to a method should be applied mutatis mutandis to a device or system capable of executing the method and/or to a non-transitory computer readable medium that stores instructions for executing the method.
Any reference in the specification to a system or device should be applied mutatis mutandis to a method that may be executed by the system, and/or may be applied mutatis mutandis to non-transitory computer readable medium that stores instructions executable by the system.
Any reference in the specification to a non-transitory computer readable medium should be applied mutatis mutandis to a device or system capable of executing instructions stored in the non-transitory computer readable medium and/or may be applied mutatis mutandis to a method for executing the instructions.
Any combination of any module or unit listed in any of the figures, any part of the specification and/or any claims may be provided.
The specification and/or drawings may refer to an image. An image is an example of a media unit. Any reference to an image may be applied mutatis mutandis to a media unit. A media unit may be an example of sensed information. Any reference to a media unit may be applied mutatis mutandis to any type of natural signal such as but not limited to signal generated by nature, signal representing human behavior, signal representing operations related to the stock market, a medical signal, financial series, geodetic signals, geophysical, chemical, molecular, textual and numerical signals, time series, and the like. Any reference to a media unit may be applied mutatis mutandis to sensed information. The sensed information may be of any kind and may be sensed by any type of sensors—such as a visual light camera, an audio sensor, a sensor that may sense infrared, radar imagery, ultrasound, electro-optics, radiography, LIDAR (light detection and ranging), etc. The sensing may include generating samples (for example, pixel, audio signals) that represent the signal that was transmitted, or otherwise reach the sensor.
The specification and/or drawings may refer to a spanning element. A spanning element may be implemented in software or hardware. Different spanning element of a certain iteration are configured to apply different mathematical functions on the input they receive. Non-limiting examples of the mathematical functions include filtering, although other functions may be applied.
The specification and/or drawings may refer to a concept structure. A concept structure may include one or more clusters. Each cluster may include signatures and related metadata. Each reference to one or more clusters may be applicable to a reference to a concept structure.
The specification and/or drawings may refer to a processor. The processor may be a processing circuitry. The processing circuitry may be implemented as a central processing unit (CPU), and/or one or more other integrated circuits such as application-specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), full-custom integrated circuits, etc., or a combination of such integrated circuits.
Any combination of any steps of any method illustrated in the specification and/or drawings may be provided.
Any combination of any subject matter of any of claims may be provided.
Any combinations of systems, units, components, processors, sensors, illustrated in the specification and/or drawings may be provided.
Any reference to an object may be applicable to a pattern. Accordingly—any reference to object detection is applicable mutatis mutandis to a pattern detection.
Low Power Generation of Signatures
The analysis of content of a media unit may be executed by generating a signature of the media unit and by comparing the signature to reference signatures. The reference signatures may be arranged in one or more concept structures or may be arranged in any other manner. The signatures may be used for object detection or for any other use.
The signature may be generated by creating a multidimensional representation of the media unit. The multidimensional representation of the media unit may have a very large number of dimensions. The high number of dimensions may guarantee that the multidimensional representation of different media units that include different objects is sparse—and that object identifiers of different objects are distant from each other—thus improving the robustness of the signatures.
The generation of the signature is executed in an iterative manner that includes multiple iterations, each iteration may include an expansion operations that is followed by a merge operation. The expansion operation of an iteration is performed by spanning elements of that iteration. By determining, per iteration, which spanning elements (of that iteration) are relevant—and reducing the power consumption of irrelevant spanning elements—a significant amount of power may be saved.
In many cases, most of the spanning elements of an iteration are irrelevant—thus after determining (by the spanning elements) their relevancy—the spanning elements that are deemed to be irrelevant may be shut down a/or enter an idle mode.
Method 5000 may start by step 5010 of receiving or generating sensed information.
The sensed information may be a media unit of multiple objects.
Step 5010 may be followed by processing the media unit by performing multiple iterations, wherein at least some of the multiple iterations comprises applying, by spanning elements of the iteration, dimension expansion process that are followed by a merge operation.
The processing may include:
-
- Step 5020 of performing a k′th iteration expansion process (k may be a variable that is used to track the number of iterations).
- Step 5030 of performing a k′th iteration merge process.
- Step 5040 of changing the value of k.
- Step 5050 of checking if all required iterations were done—if so proceeding to step 5060 of completing the generation of the signature. Else—jumping to step 5020.
The output of step 5020 is a k′th iteration expansion results 5120.
The output of step 5030 is a k′th iteration merge results 5130.
For each iteration (except the first iteration)—the merge result of the previous iteration is an input to the current iteration expansion process.
At least some of the K iterations involve selectively reducing the power consumption of some spanning elements (during step 5020) that are deemed to be irrelevant.
The image 6001 is virtually segments to segments 6000(i,k). The segments may be of the same shape and size but this is not necessarily so.
Outcome 6013 may be a tensor that includes a vector of values per each segment of the media unit. One or more objects may appear in a certain segment. For each object—an object identifier (of the signature) points to locations of significant values, within a certain vector associated with the certain segment.
For example—a top left segment (6001(1,1)) of the image may be represented in the outcome 6013 by a vector V(1,1) 6017(1,1) that has multiple values. The number of values per vector may exceed 100, 200, 500, 1000, and the like.
The significant values (for example—more than 10, 20, 30, 40 values, and/or more than 0.1%, 0.2%. 0.5%, 1%, 5% of all values of the vector and the like) may be selected. The significant values may have the values—but may eb selected in any other manner.
The image signature 6027 includes five indexes for the retrieval of the five significant values—first till fifth identifiers ID1-ID5 are indexes for retrieving the first till fifth significant values.
The k′th iteration expansion process start by receiving the merge results 5060′ of a previous iteration.
The merge results of a previous iteration may include values are indicative of previous expansion processes—for example—may include values that are indicative of relevant spanning elements from a previous expansion operation, values indicative of relevant regions of interest in a multidimensional representation of the merge results of a previous iteration.
The merge results (of the previous iteration) are fed to spanning elements such as spanning elements 5061(1)-5061(J).
Each spanning element is associated with a unique set of values. The set may include one or more values. The spanning elements apply different functions that may be orthogonal to each other. Using non-orthogonal functions may increase the number of spanning elements—but this increment may be tolerable.
The spanning elements may apply functions that are decorrelated to each other—even if not orthogonal to each other.
The spanning elements may be associated with different combinations of object identifiers that may “cover” multiple possible media units. Candidates for combinations of object identifiers may be selected in various manners—for example based on their occurrence in various images (such as test images) randomly, pseudo randomly, according to some rules and the like. Out of these candidates the combinations may be selected to be decorrelated, to cover said multiple possible media units and/or in a manner that certain objects are mapped to the same spanning elements.
Each spanning element compares the values of the merge results to the unique set (associated with the spanning element) and if there is a match—then the spanning element is deemed to be relevant. If so—the spanning element completes the expansion operation.
If there is no match—the spanning element is deemed to be irrelevant and enters a low power mode. The low power mode may also be referred to as an idle mode, a standby mode, and the like. The low power mode is termed low power because the power consumption of an irrelevant spanning element is lower than the power consumption of a relevant spanning element.
In
Each relevant spanning element may perform a spanning operation that includes assigning an output value that is indicative of an identity of the relevant spanning elements of the iteration. The output value may also be indicative of identities of previous relevant spanning elements (from previous iterations).
For example—assuming that spanning element number fifty is relevant and is associated with a unique set of values of eight and four—then the output value may reflect the numbers fifty, four and eight—for example one thousand multiplied by (fifty+forty) plus forty. Any other mapping function may be applied.
-
- Checking if the merge results are relevant to the spanning element (step 5091).
- If-so—completing the spanning operation (step 5093).
- If not—entering an idle state (step 5092).
A merge operation may include finding regions of interest. The regions of interest are regions within a multidimensional representation of the sensed information. A region of interest may exhibit a more significant response (for example a stronger, higher intensity response).
The merge operation (executed during a k′th iteration merge operation) may include at least one of the following:
-
- Step 5031 of searching for overlaps between regions of interest (of the k′th iteration expansion operation results) and define regions of interest that are related to the overlaps.
- Step 5032 of determining to drop one or more region of interest, and dropping according to the determination.
- Step 5033 of searching for relationships between regions of interest (of the k′th iteration expansion operation results) and define regions of interest that are related to the relationship.
- Step 5034 of searching for proximate regions of interest (of the k′th iteration expansion operation results) and define regions of interest that are related to the proximity. Proximate may be a distance that is a certain fraction (for example less than 1%) of the multi-dimensional space, may be a certain fraction of at least one of the regions of interest that are tested for proximity.
- Step 5035 of searching for relationships between regions of interest (of the k′th iteration expansion operation results) and define regions of interest that are related to the relationship.
- Step 5036 of merging and/or dropping k′th iteration regions of interest based on shape information related to shape of the k′th iteration regions of interest.
The same merge operations may applied in different iterations.
Alternatively, different merge operations may be executed during different iterations.
The hybrid process is hybrid in the sense that some expansion and merge operations are executed by a convolutional neural network (CNN) and some expansion and merge operations (denoted additional iterations of expansion and merge) are not executed by the CNN—but rather by a process that may include determining a relevancy of spanning elements and entering irrelevant spanning elements to a low power mode.
In
The output of these layers provided information about image properties. The image properties may not amount to object detection. Image properties may include location of edges, properties of curves, and the like.
The CNN may include additional layers (for example third till N′th layer 6010(N)) that may provide a CNN output 6018 that may include object detection information. It should be noted that the additional layers may not be included.
It should be noted that executing the entire signature generation process by a hardware CNN of fixed connectivity may have a higher power consumption—as the CNN will not be able to reduce the power consumption of irrelevant nodes.
In
The first expansion operation involves filtering the input image by a first filtering operation 6031 to provide first regions of interest (denoted 1) in a first filtered image 6031′.
The first expansion operation also involves filtering the input image by a second filtering operation 6032 to provide first regions of interest (denoted 2) in a second filtered image 6032′,
The merge operation includes merging the two images by overlaying the first filtered image on the second filtered image to provide regions of interest 1, 2, 12 and 21. Region of interest 12 is an overlap area shared by a certain region of interest 1 and a certain region of interest 2. Region of interest 21 is a union of another region of interest 1 and another region of interest 2.
Method 5200 may include the following sequence of steps:
-
- Step 5210 of receiving or generating an image.
- Step 5220 of performing a first iteration expansion operation (which is an expansion operation that is executed during a first iteration)
- Step 5230 of performing a first iteration merge operation.
- Step 5240 of amending index k (k is an iteration counter). In
FIG. 7 in incremented by one—this is only an example of how the number of iterations are tracked. - Step 5260 of performing a k′th iteration expansion operation on the (k−1)′th iteration merge results.
- Step 5270 of performing a k′th iteration merge operation (on the k′th iteration expansion operation results.
- Step 5280 of changing the value of index k.
- Step 5290 of checking if all iteration ended (k reached its final value—for example K).
If no—there are still iterations to be executed—jumping from step 5290 to step 5260.
If yes—jumping to step 5060 of completing the generation of the signature. This may include, for example, selecting significant attributes, determining retrieval information (for example indexes) that point to the selected significant attributes.
Step 5220 may include:
-
- Step 5222 of generating multiple representations of the image within a multi-dimensional space of f(1) dimensions. The expansion operation of step 5220 generates a first iteration multidimensional representation of the first image. The number of dimensions of this first iteration multidimensional representation is denoted f(1).
- Step 5224 of assigning a unique index for each region of interest within the multiple representations. For example, referring to
FIG. 6 —th indexes 1 and indexes 2 are assigned to regions of interests generated during the first iteration expansion operations 6031 and 6032.
Step 5230 may include:
-
- Step 5232 of searching for relationships between regions of interest and define regions of interest that are related to the relationships. For example—union or intersection illustrate din
FIG. 6 . - Step 5234 of assigning a unique index for each region of interest within the multiple representations. For example—referring to
FIG. 6 —indexes 1, 2, 12 and 21.
- Step 5232 of searching for relationships between regions of interest and define regions of interest that are related to the relationships. For example—union or intersection illustrate din
Step 5260 may include:
-
- Step 5262 of generating multiple representations of the merge results of the (k−1)′th iteration within a multi-dimensional space of f(k) dimensions. The expansion operation of step 5260 generates a k′th iteration multidimensional representation of the first image. The number of dimensions of this kth iteration multidimensional representation is denoted f(k).
- Step 5264 of assigning a unique index for each region of interest within the multiple representations.
Step 5270 may include
-
- Step 5272 of searching for relationships between regions of interest and define regions of interest that are related to the relationships.
- Step 5274 of Assigning a unique index for each region of interest within the multiple representations.
Thus—step 5232 is replaced by step 5232′ of searching for overlaps between regions of interest and define regions of interest that are related to the overlaps.
Step 5272 is replaced by step 5272′ of searching for overlaps between regions of interest and define regions of interest that are related to the overlaps.
Method 7000 starts by step 7010 of receiving or generating a media unit of multiple objects.
Step 7010 may be followed by step 7012 of processing the media unit by performing multiple iterations, wherein at least some of the multiple iterations comprises applying, by spanning elements of the iteration, dimension expansion process that are followed by a merge operation.
The applying of the dimension expansion process of an iteration may include (a) determining a relevancy of the spanning elements of the iteration; and (b) completing the dimension expansion process by relevant spanning elements of the iteration and reducing a power consumption of irrelevant spanning elements until, at least, a completion of the applying of the dimension expansion process.
The identifiers may be retrieval information for retrieving the significant portions.
The at least some of the multiple iterations may be a majority of the multiple iterations.
The output of the multiple iteration may include multiple property attributes for each segment out of multiple segments of the media unit; and wherein the significant portions of an output of the multiple iterations may include more impactful property attributes.
The first iteration of the multiple iteration may include applying the dimension expansion process by applying different filters on the media unit.
The at least some of the multiple iteration exclude at least a first iteration of the multiple iterations. See, for example,
The determining the relevancy of the spanning elements of the iteration may be based on at least some identities of relevant spanning elements of at least one previous iteration.
The determining the relevancy of the spanning elements of the iteration may be based on at least some identities of relevant spanning elements of at least one previous iteration that preceded the iteration.
The determining the relevancy of the spanning elements of the iteration may be based on properties of the media unit.
The determining the relevancy of the spanning elements of the iteration may be performed by the spanning elements of the iteration.
Method 7000 may include a neural network processing operation that may be executed by one or more layers of a neural network and does not belong to the at least some of the multiple iterations. See, for example,
The at least one iteration may be executed without reducing power consumption of irrelevant neurons of the one or more layers.
The one or more layers may output information about properties of the media unit, wherein the information differs from a recognition of the multiple objects.
The applying, by spanning elements of an iteration that differs from the first iteration, the dimension expansion process may include assigning output values that may be indicative of an identity of the relevant spanning elements of the iteration. See, for example,
The applying, by spanning elements of an iteration that differs from the first iteration, the dimension expansion process may include assigning output values that may be indicative a history of dimension expansion processes until the iteration that differs from the first iteration.
The each spanning element may be associated with a subset of reference identifiers. The determining of the relevancy of each spanning elements of the iteration may be based a relationship between the subset of the reference identifiers of the spanning element and an output of a last merge operation before the iteration.
The output of a dimension expansion process of an iteration may be a multidimensional representation of the media unit that may include media unit regions of interest that may be associated with one or more expansion processes that generated the regions of interest.
The merge operation of the iteration may include selecting a subgroup of media unit regions of interest based on a spatial relationship between the subgroup of multidimensional regions of interest. See, for example,
Method 7000 may include applying a merge function on the subgroup of multidimensional regions of interest. See, for example,
Method 7000 may include applying an intersection function on the subgroup of multidimensional regions of interest. See, for example,
The merge operation of the iteration may be based on an actual size of one or more multidimensional regions of interest.
The merge operation of the iteration may be based on relationship between sizes of the multidimensional regions of interest. For example—larger multidimensional regions of interest may be maintained while smaller multidimensional regions of interest may be ignored of.
The merge operation of the iteration may be based on changes of the media unit regions of interest during at least the iteration and one or more previous iteration.
Step 7012 may be followed by step 7014 of determining identifiers that are associated with significant portions of an output of the multiple iterations.
Step 7014 may be followed by step 7016 of providing a signature that comprises the identifiers and represents the multiple objects.
Localization and Segmentation
Any of the mentioned above signature generation method provides a signature that does not explicitly includes accurate shape information. This adds to the robustness of the signature to shape related inaccuracies or to other shape related parameters.
The signature includes identifiers for identifying media regions of interest.
Each media region of interest may represent an object (for example a vehicle, a pedestrian, a road element, a human made structure, wearables, shoes, a natural element such as a tree, the sky, the sun, and the like) or a part of an object (for example—in the case of the pedestrian—a neck, a head, an arm, a leg, a thigh, a hip, a foot, an upper arm, a forearm, a wrist, and a hand). It should be noted that for object detection purposes a part of an object may be regarded as an object.
The exact shape of the object may be of interest.
Method 7002 may include a sequence of steps 7020, 7022, 7024 and 7026.
Step 7020 may include receiving or generating the media unit.
Step 7022 may include processing the media unit by performing multiple iterations, wherein at least some of the multiple iterations comprises applying, by spanning elements of the iteration, dimension expansion process that are followed by a merge operation.
Step 7024 may include selecting, based on an output of the multiple iterations, media unit regions of interest that contributed to the output of the multiple iterations.
Step 7026 may include providing a hybrid representation, wherein the hybrid representation may include (a) shape information regarding shapes of the media unit regions of interest, and (b) a media unit signature that includes identifiers that identify the media unit regions of interest.
Step 7024 may include selecting the media regions of interest per segment out of multiple segments of the media unit. See, for example,
Step 7026 may include step 7027 of generating the shape information.
The shape information may include polygons that represent shapes that substantially bound the media unit regions of interest. These polygons may be of a high degree.
In order to save storage space, the method may include step 7028 of compressing the shape information of the media unit to provide compressed shape information of the media unit.
Method 5002 may start by step 5011 of receiving or generating a media unit.
Step 5011 may be followed by processing the media unit by performing multiple iterations, wherein at least some of the multiple iterations comprises applying, by spanning elements of the iteration, dimension expansion process that are followed by a merge operation.
The processing may be followed by steps 5060 and 5062.
The processing may include steps 5020, 5030, 5040 and 5050.
Step 5020 may include performing a k′th iteration expansion process (k may be a variable that is used to track the number of iterations).
Step 5030 may include performing a k′th iteration merge process.
Step 5040 may include changing the value of k.
Step 5050 may include checking if all required iterations were done—if so proceeding to steps 5060 and 5062. Else—jumping to step 5020.
The output of step 5020 is a k′th iteration expansion result.
The output of step 5030 is a k′th iteration merge result.
For each iteration (except the first iteration)—the merge result of the previous iteration is an input to the current iteration expansion process.
Step 5060 may include completing the generation of the signature.
Step 5062 may include generating shape information regarding shapes of media unit regions of interest. The signature and the shape information provide a hybrid representation of the media unit.
The combination of steps 5060 and 5062 amounts to a providing a hybrid representation, wherein the hybrid representation may include (a) shape information regarding shapes of the media unit regions of interest, and (b) a media unit signature that includes identifiers that identify the media unit regions of interest.
Method 5200 may include the following sequence of steps:
-
- Step 5210 of receiving or generating an image.
- Step 5230 of performing a first iteration expansion operation (which is an expansion operation that is executed during a first iteration)
- Step 5240 of performing a first iteration merge operation.
- Step 5240 of amending index k (k is an iteration counter). In
FIG. 1L in incremented by one—this is only an example of how the number of iterations are tracked. - Step 5260 of performing a k′th iteration expansion operation on the (k−1)′th iteration merge results.
- Step 5270 of Performing a k′th iteration merge operation (on the k′th iteration expansion operation results.
- Step 5280 of changing the value of index k.
- Step 5290 of checking if all iteration ended (k reached its final value—for example K).
If no—there are still iterations to be executed—jumping from step 5290 to step 5260.
If yes—jumping to step 5060.
Step 5060 may include completing the generation of the signature. This may include, for example, selecting significant attributes, determining retrieval information (for example indexes) that point to the selected significant attributes.
Step 5062 may include generating shape information regarding shapes of media unit regions of interest. The signature and the shape information provide a hybrid representation of the media unit.
The combination of steps 5060 and 5062 amounts to a providing a hybrid representation, wherein the hybrid representation may include (a) shape information regarding shapes of the media unit regions of interest, and (b) a media unit signature that includes identifiers that identify the media unit regions of interest.
Step 5220 may include:
-
- Step 5222 of generating multiple representations of the image within a multi-dimensional space of f(k) dimensions.
- Step 5224 of assigning a unique index for each region of interest within the multiple representations. (for example, referring to
FIG. 1F —indexes 1 and indexes 2 following first iteration expansion operations 6031 and 6032.
Step 5230 may include
-
- Step 5226 of searching for relationships between regions of interest and define regions of interest that are related to the relationships. For example—union or intersection illustrated in
FIG. 1F . - Step 5228 of assigning a unique index for each region of interest within the multiple representations. For example—referring to
FIG. 1F —indexes 1, 2, 12 and 21.
- Step 5226 of searching for relationships between regions of interest and define regions of interest that are related to the relationships. For example—union or intersection illustrated in
Step 5260 may include:
-
- Step 5262 of generating multiple representations of the merge results of the (k−1)′th iteration within a multi-dimensional space of f(k) dimensions. The expansion operation of step 5260 generates a k′th iteration multidimensional representation of the first image. The number of dimensions of this kth iteration multidimensional representation is denoted f(k).
- Step 5264 of assigning a unique index for each region of interest within the multiple representations.
Step 5270 may include
-
- Step 5272 of searching for relationships between regions of interest and define regions of interest that are related to the relationships.
- Step 5274 of assigning a unique index for each region of interest within the multiple representations.
Method 5200 may include the following sequence of steps:
-
- Step 5210 of receiving or generating an image.
- Step 5230 of performing a first iteration expansion operation (which is an expansion operation that is executed during a first iteration)
- Step 5240 of performing a first iteration merge operation.
- Step 5240 of amending index k (k is an iteration counter). In
FIG. 1M in incremented by one—this is only an example of how the number of iterations are tracked. - Step 5260 of performing a k′th iteration expansion operation on the (k−1)′th iteration merge results.
- Step 5270 of performing a k′th iteration merge operation (on the k′th iteration expansion operation results.
- Step 5280 of changing the value of index k.
- Step 5290 of checking if all iteration ended (k reached its final value—for example K).
If no—there are still iterations to be executed—jumping from step 5290 to step 5260.
If yes—jumping to steps 5060 and 5062.
Step 5060 may include completing the generation of the signature. This may include, for example, selecting significant attributes, determining retrieval information (for example indexes) that point to the selected significant attributes.
Step 5062 may include generating shape information regarding shapes of media unit regions of interest. The signature and the shape information provide a hybrid representation of the media unit.
The combination of steps 5060 and 5062 amounts to a providing a hybrid representation, wherein the hybrid representation may include (a) shape information regarding shapes of the media unit regions of interest, and (b) a media unit signature that includes identifiers that identify the media unit regions of interest.
Step 5220 may include:
-
- Step 5221 of filtering the image with multiple filters that are orthogonal to each other to provide multiple filtered images that are representations of the image in a multi-dimensional space of f(1) dimensions. The expansion operation of step 5220 generates a first iteration multidimensional representation of the first image. The number of filters is denoted f(1).
- Step 5224 of assigning a unique index for each region of interest within the multiple representations. (for example, referring to
FIG. 1F —indexes 1 and indexes 2 following first iteration expansion operations 6031 and 6032.
Step 5230 may include
-
- Step 5226 of searching for relationships between regions of interest and define regions of interest that are related to the relationships. For example—union or intersection illustrated in
FIG. 1F . - Step 5228 of assigning a unique index for each region of interest within the multiple representations. For example—referring to
FIG. 1F —indexes 1, 2, 12 and 21.
- Step 5226 of searching for relationships between regions of interest and define regions of interest that are related to the relationships. For example—union or intersection illustrated in
Step 5260 may include:
-
- Step 5262 of generating multiple representations of the merge results of the (k−1)′th iteration within a multi-dimensional space of f(k) dimensions. The expansion operation of step 5260 generates a k′th iteration multidimensional representation of the first image. The number of dimensions of this kth iteration multidimensional representation is denoted f(k).
- Step 5264 of assigning a unique index for each region of interest within the multiple representations.
Step 5270 may include
-
- Step 5272 of searching for relationships between regions of interest and define regions of interest that are related to the relationships.
- Step 5274 of assigning a unique index for each region of interest within the multiple representations.
The filters may be orthogonal may be non-orthogonal—for example be decorrelated. Using non-orthogonal filters may increase the number of filters—but this increment may be tolerable.
Object Detection Using Compressed Shape Information
Object detection may include comparing a signature of an input image to signatures of one or more cluster structures in order to find one or more cluster structures that include one or more matching signatures that match the signature of the input image.
The number of input images that are compared to the cluster structures may well exceed the number of signatures of the cluster structures. For example—thousands, tens of thousands, hundreds of thousands (and even more) of input signature may be compared to much less cluster structure signatures. The ratio between the number of input images to the aggregate number of signatures of all the cluster structures may exceed ten, one hundred, one thousand, and the like.
In order to save computational resources, the shape information of the input images may be compressed.
On the other hand—the shape information of signatures that belong to the cluster structures may be uncompressed—and of higher accuracy than those of the compressed shape information.
When the higher quality is not required—the shape information of the cluster signature may also be compressed.
Compression of the shape information of cluster signatures may be based on a priority of the cluster signature, a popularity of matches to the cluster signatures, and the like.
The shape information related to an input image that matches one or more of the cluster structures may be calculated based on shape information related to matching signatures.
For example—a shape information regarding a certain identifier within the signature of the input image may be determined based on shape information related to the certain identifiers within the matching signatures.
Any operation on the shape information related to the certain identifiers within the matching signatures may be applied in order to determine the (higher accuracy) shape information of a region of interest of the input image identified by the certain identifier.
For example—the shapes may be virtually overlaid on each other and the population per pixel may define the shape.
For example—only pixels that appear in at least a majority of the overlaid shaped should be regarded as belonging to the region of interest.
Other operations may include smoothing the overlaid shapes, selecting pixels that appear in all overlaid shapes.
The compressed shape information may be ignored of or be taken into account.
Method 7003 may include a sequence of steps 7030, 7032 and 7034.
Step 7030 may include receiving or generating a hybrid representation of a media unit. The hybrid representation includes compressed shape information.
Step 7032 may include comparing the media unit signature of the media unit to signatures of multiple concept structures to find a matching concept structure that has at least one matching signature that matches to the media unit signature.
Step 7034 may include calculating higher accuracy shape information that is related to regions of interest of the media unit, wherein the higher accuracy shape information is of higher accuracy than the compressed shape information of the media unit, wherein the calculating is based on shape information associated with at least some of the matching signatures.
Step 7034 may include at least one out of:
-
- Determining shapes of the media unit regions of interest using the higher accuracy shape information.
- For each media unit region of interest, virtually overlaying shapes of corresponding media units of interest of at least some of the matching signatures.
It is assumed that there are multiple (M) cluster structures 4974(1)-4974(M). Each cluster structure includes cluster signatures, metadata regarding the cluster signatures, and shape information regarding the regions of interest identified by identifiers of the cluster signatures.
For example—first cluster structure 4974(1) includes multiple (N1) signatures (referred to as cluster signatures CS) CS(1,1)-CS(1,N1) 4975(1,1)-4975(1,N1), metadata 4976(1), and shape information (Shapeinfo 4977(1)) regarding shapes of regions of interest associated with identifiers of the CSs.
Yet for another example—M′th cluster structure 4974(M) includes multiple (N2) signatures (referred to as cluster signatures CS) CS(M,1)-CS(M,N2) 4975(M,1)-4975(M,N2), metadata 4976(M), and shape information (Shapeinfo 4977(M)) regarding shapes of regions of interest associated with identifiers of the CSs.
The number of signatures per concept structure may change over time—for example due to cluster reduction attempts during which a CS is removed from the structure to provide a reduced cluster structure, the reduced structure structure is checked to determine that the reduced cluster signature may still identify objects that were associated with the (non-reduced) cluster signature—and if so the signature may be reduced from the cluster signature.
The signatures of each cluster structures are associated to each other, wherein the association may be based on similarity of signatures and/or based on association between metadata of the signatures.
Assuming that each cluster structure is associated with a unique object—then objects of a media unit may be identified by finding cluster structures that are associated with said objects. The finding of the matching cluster structures may include comparing a signature of the media unit to signatures of the cluster structures—and searching for one or more matching signature out of the cluster signatures.
In
The media unit signature 4972 is compared to the signatures of the M cluster structures—from CS(1,1) 4975(1,1) till CS(M,N2) 4975(M,N2).
We assume that one or more cluster structures are matching cluster structures.
Once the matching cluster structures are found the method proceeds by generating shape information that is of higher accuracy then the compressed shape information.
The generation of the shape information is done per identifier.
For each j that ranges between 1 and J (J is the number of identifiers per the media unit signature 4972) the method may perform the steps of:
-
- Find (step 4978(j)) the shape information of the j′th identifier of each matching signature—or of each signature of the matching cluster structure.
- Generate (step 4979(j)) a higher accuracy shape information of the j′th identifier.
For example—assuming that the matching signatures include CS(1,1) 2975(1,1), CS(2,5) 2975(2,5), CS(7,3) 2975(7,3) and CS(15,2) 2975(15,2), and that the j′th identifier is included in CS(1,1) 2975(1,1),CS(7,3) 2975(7,3) and CS(15,2) 2975(15,2)—then the shape information of the j′th identifier of the media unit is determined based on the shape information associated with CS(1,1) 2975(1,1),CS(7,3) 2975(7,3) and CS(15,2) 2975(15,2).
The shapes of the four regions of interest 8001, 8002, 8003 and 8004 are four polygons. Accurate shape information regarding the shapes of these regions of interest may be generated during the generation of signature 8010.
The hybrid representation of the media unit, after compression represent an media unit with simplified regions of interest 8001′, 8002′, 8003′ and 8004′—as shown in
Scale Based Bootstrap
Objects may appear in an image at different scales. Scale invariant object detection may improve the reliability and repeatability of the object detection and may also use fewer number of cluster structures—thus reduced memory resources and also lower computational resources required to maintain fewer cluster structures.
Method 8020 may include a first sequence of steps that may include step 8022, 8024, 8026 and 8028.
Step 8022 may include receiving or generating a first image in which an object appears in a first scale and a second image in which the object appears in a second scale that differs from the first scale.
Step 8024 may include generating a first image signature and a second image signature.
The first image signature includes a first group of at least one certain first image identifier that identifies at least a part of the object. See, for example image 8000′ of
The second image signature includes a second group of certain second image identifiers that identify different parts of the object.
See, for example image 8000 of
The second group is larger than first group—as the second group has more members than the first group.
Step 8026 may include linking between the at least one certain first image identifier and the certain second image identifiers.
Step 8026 may include linking between the first image signature, the second image signature and the object.
Step 8026 may include adding the first signature and the second signature to a certain concept structure that is associated with the object. For example, referring to
Step 8028 may include determining whether an input image includes the object based, at least in part, on the linking. The input image differs from the first and second images.
The determining may include determining that the input image includes the object when a signature of the input image includes the at least one certain first image identifier or the certain second image identifiers.
The determining may include determining that the input image includes the object when the signature of the input image includes only a part of the at least one certain first image identifier or only a part of the certain second image identifiers.
The linking may be performed for more than two images in which the object appears in more than two scales.
For example, see
In first image 8051 the person is included in a single region of interest 8061 and the signature 8051′ of first image 8051 includes an identifier ID61 that identifies the single region of interest—identifies the person.
In second image 8052 the upper part of the person is included in region of interest 8068, the lower part of the person is included in region of interest 8069 and the signature 8052′ of second image 8052 includes identifiers ID68 and ID69 that identify regions of interest 8068 and 8069 respectively.
In third image 8053 the eyes of the person are included in region of interest 8062, the mouth of the person is included in region of interest 8063, the head of the person appears in region of interest 8064, the neck and arms of the person appear in region of interest 8065, the middle part of the person appears in region of interest 8066, and the lower part of the person appears in region of interest 8067. Signature 8053′ of third image 8053 includes identifiers ID62, ID63, ID64, ID65, ID55 and ID67 that identify regions of interest 8062-8067 respectively.
Method 8020 may link signatures 8051′, 8052′ and 8053′ to each other. For example—these signatures may be included in the same cluster structure.
Method 8020 may link (i) ID61, (ii) signatures ID68 and ID69, and (ii) signature ID62, ID63, ID64, ID65, ID66 and ID67.
Method 8030 may include the steps of method 8020 or may be preceded by steps 8022, 8024 and 8026.
Method 8030 may include a sequence of steps 8032, 8034, 8036 and 8038.
Step 8032 may include receiving or generating an input image.
Step 8034 may include generating a signature of the input image.
Step 8036 may include comparing the signature of the input image to signatures of a certain concept structure. The certain concept structure may be generated by method 8020.
Step 8038 may include determining that the input image comprises the object when at least one of the signatures of the certain concept structure matches the signature of the input image.
Method 8040 may include the steps of method 8020 or may be preceded by steps 8022, 8024 and 8026.
Method 8040 may include a sequence of steps 8041, 8043, 8045, 8047 and 8049.
Step 8041 may include receiving or generating an input image.
Step 8043 may include generating a signature of the input image, the signature of the input image comprises only some of the certain second image identifiers; wherein the input image of the second scale.
Step 8045 may include changing a scale of the input image to the first scale to a provide an amended input image.
Step 8047 may include generating a signature of the amended input image.
Step 8049 may include verifying that the input image comprises the object when the signature of the amended input image comprises the at least one certain first image identifier.
Method 8050 may include the steps of method 8020 or may be preceded by steps 8022, 8024 and 8026.
Method 8050 may include a sequence of steps 8052, 8054, 8056 and 8058.
Step 8052 may include receiving or generating an input image.
Step 8054 may include generating a signature of the input image.
Step 8056 may include searching in the signature of the input image for at least one of (a) the at least one certain first image identifier, and (b) the certain second image identifiers.
Step 8058 may include determining that the input image comprises the object when the signature of the input image comprises the at least one of (a) the at least one certain first image identifier, and (b) the certain second image identifiers.
It should be noted that step 8056 may include searching in the signature of the input image for at least one of (a) one or more certain first image identifier of the at least one certain first image identifier, and (b) at least one certain second image identifier of the certain second image identifiers.
It should be noted that step 8058 may include determining that the input image includes the object when the signature of the input image comprises the at least one of (a) one or more certain first image identifier of the at least one certain first image identifier, and (b) the at least one certain second image identifier.
Movement Based Bootstrapping
A single object may include multiple parts that are identified by different identifiers of a signature of the image. In cases such as unsupervised learning, it may be beneficial to link the multiple object parts to each other without receiving prior knowledge regarding their inclusion in the object.
Additionally or alternatively, the linking can be done in order to verify a previous linking between the multiple object parts.
Method 8070 is for movement based object detection.
Method 8070 may include a sequence of steps 8071, 8073, 8075, 8077, 8078 and 8079.
Step 8071 may include receiving or generating a video stream that includes a sequence of images.
Step 8073 may include generating image signatures of the images. Each image is associated with an image signature that comprises identifiers. Each identifier identifiers a region of interest within the image.
Step 8075 may include generating movement information indicative of movements of the regions of interest within consecutive images of the sequence of images. Step 8075 may be preceded by or may include generating or receiving location information indicative of a location of each region of interest within each image. The generating of the movement information is based on the location information.
Step 8077 may include searching, based on the movement information, for a first group of regions of interest that follow a first movement. Different first regions of interest are associated with different parts of an object.
Step 8078 may include linking between first identifiers that identify the first group of regions of interest.
Step 8079 may include linking between first image signatures that include the first linked identifiers.
Step 8079 may include adding the first image signatures to a first concept structure, the first concept structure is associated with the first image.
Step 8079 may be followed by determining whether an input image includes the object based, at least in part, on the linking
An example of various steps of method 8070 is illustrated in
First image 8091 illustrates a gate 8089′ that is located in region of interest 8089 and a person that faces the gate. Various parts of the person are located within regions of interest 8081, 8082, 8083, 8084 and 8085.
The first image signature 8091′ includes identifiers ID81, ID82, ID83, ID84, ID85 and ID89 that identify regions of interest 8081, 8082, 8083, 8084, 8085 and 8089 respectively.
The first image location information 8091″ includes the locations L81, L82, L83, L84, L85 and L89 of regions of interest 8081, 8082, 8083, 8084, 8085 and 8089 respectively. A location of a region of interest may include a location of the center of the region of interest, the location of the borders of the region of interest or any location information that may define the location of the region of interest or a part of the region of interest.
Second image 8092 illustrates a gate that is located in region of interest 8089 and a person that faces the gate. Various parts of the person are located within regions of interest 8081, 8082, 8083, 8084 and 8085. Second image also includes a pole that is located within region of interest 8086. In the first image that the pole was concealed by the person.
The second image signature 8092′ includes identifiers ID81, ID82, ID83, ID84, ID85, ID86, and ID89 that identify regions of interest 8081, 8082, 8083, 8084, 8085, 8086 and 8089 respectively.
The second image location information 8092″ includes the locations L81, L82, L83, L84, L85, L86 and L89 of regions of interest 8081, 8082, 8083, 8084, 8085, 8086 and 8089 respectively.
Third image 8093 illustrates a gate that is located in region of interest 8089 and a person that faces the gate. Various parts of the person are located within regions of interest 8081, 8082, 8083, 8084 and 8085. Third image also includes a pole that is located within region of interest 8086, and a balloon that is located within region of interest 8087.
The third image signature 8093′ includes identifiers ID81, ID82, ID83, ID84, ID85, ID86, ID87 and ID89 that identify regions of interest 8081, 8082, 8083, 8084, 8085, 8086, 8087 and 8089 respectively.
The third image location information 8093″ includes the locations L81, L82, L83, L84, L85, L86, L87 and L89 of regions of interest 8081, 8082, 8083, 8084, 8085, 8086, 8086 and 8089 respectively.
The motion of the various regions of interest may be calculated by comparing the location information related to different images. The movement information may take into account the different in the acquisition time of the images.
The comparison shows that regions of interest 8081, 8082, 8083, 8084, 8085 move together and thus they should be linked to each other—and it may be assumed that they all belong to the same object.
Method 8100 may include the steps of method 8070 or may be preceded by steps 8071, 8073, 8075, 8077 and 8078.
Method 8100 may include the following sequence of steps:
-
- Step 8102 of receiving or generating an input image.
- Step 8104 of generating a signature of the input image.
- Step 8106 of comparing the signature of the input image to signatures of a first concept structure. The first concept structure includes first identifiers that were linked to each other based on movements of first regions of interest that are identified by the first identifiers.
- Step 8108 of determining that the input image includes a first object when at least one of the signatures of the first concept structure matches the signature of the input image.
Method 8110 may include the steps of method 8070 or may be preceded by steps 8071, 8073, 8075, 8077 and 8078.
Method 8110 may include the following sequence of steps:
-
- Step 8112 of receiving or generating an input image.
- Step 8114 of generating a signature of the input image.
- Step 8116 of searching in the signature of the input image for at least one of the first identifiers.
- Step 8118 of determining that the input image comprises the object when the signature of the input image comprises at least one of the first identifiers.
Object Detection that is Robust to Angle of Acquisition
Object detection may benefit from being robust to the angle of acquisition—to the angle between the optical axis of an image sensor and a certain part of the object. This allows the detection process to be more reliable, use fewer different clusters (may not require multiple clusters for identifying the same object from different images).
-
- Step 8122 of receiving or generating images of objects taken from different angles.
- Step 8124 of finding images of objects taken from different angles that are close to each other. Close enough may be less than 1, 5, 10, 15 and 20 degrees—but the closeness may be better reflected by the reception of substantially the same signature.
- Step 8126 of linking between the images of similar signatures. This may include searching for local similarities. The similarities are local in the sense that they are calculated per a subset of signatures. For example—assuming that the similarity is determined per two images—then a first signature may be linked to a second signature that is similar to the first image. A third signature may be linked to the second image based on the similarity between the second and third signatures—and even regardless of the relationship between the first and third signatures.
Step 8126 may include generating a concept data structure that includes the similar signatures.
This so-called local or sliding window approach, in addition to the acquisition of enough images (that will statistically provide a large angular coverage) will enable to generate a concept structure that include signatures of an object taken at multiple directions.
Signature Tailored Matching Threshold
Object detection may be implemented by (a) receiving or generating concept structures that include signatures of media units and related metadata, (b) receiving a new media unit, generating a new media unit signature, and (c) comparing the new media unit signature to the concept signatures of the concept structures.
The comparison may include comparing new media unit signature identifiers (identifiers of objects that appear in the new media unit) to concept signature identifiers and determining, based on a signature matching criteria whether the new media unit signature matches a concept signature. If such a match is found then the new media unit is regarded as including the object associated with that concept structure.
It was found that by applying an adjustable signature matching criteria, the matching process may be highly effective and may adapt itself to the statistics of appearance of identifiers in different scenarios. For example—a match may be obtained when a relatively rear but highly distinguishing identifier appears in the new media unit signature and in a cluster signature, but a mismatch may be declared when multiple common and slightly distinguishing identifiers appear in the new media unit signature and in a cluster signature.
Method 8200 may include:
-
- Step 8210 of receiving an input image.
- Step 8212 of generating a signature of the input image.
- Step 8214 of comparing the signature of the input image to signatures of a concept structure.
- Step 8216 of determining whether the signature of the input image matches any of the signatures of the concept structure based on signature matching criteria, wherein each signature of the concept structure is associated within a signature matching criterion that is determined based on an object detection parameter of the signature.
- Step 8218 of concluding that the input image comprises an object associated with the concept structure based on an outcome of the determining.
The signature matching criteria may be a minimal number of matching identifiers that indicate of a match. For example—assuming a signature that include few tens of identifiers, the minimal number may vary between a single identifier to all of the identifiers of the signature.
It should be noted that an input image may include multiple objects and that an signature of the input image may match multiple cluster structures. Method 8200 is applicable to all of the matching processes—and that the signature matching criteria may be set for each signature of each cluster structure.
Step 8210 may be preceded by step 8202 of determining each signature matching criterion by evaluating object detection capabilities of the signature under different signature matching criteria.
Step 8202 may include:
-
- Step 8203 of receiving or generating signatures of a group of test images.
- Step 8204 of calculating the object detection capability of the signature, for each signature matching criterion of the different signature matching criteria.
- Step 8206 of selecting the signature matching criterion based on the object detection capabilities of the signature under the different signature matching criteria.
The object detection capability may reflect a percent of signatures of the group of test images that match the signature.
The selecting of the signature matching criterion comprises selecting the signature matching criterion that once applied results in a percent of signatures of the group of test images that match the signature that is closets to a predefined desired percent of signatures of the group of test images that match the signature.
The object detection capability may reflect a significant change in the percent of signatures of the group of test images that match the signature. For example—assuming, that the signature matching criteria is a minimal number of matching identifiers and that changing the value of the minimal numbers may change the percentage of matching test images. A substantial change in the percentage (for example a change of more than 10, 20, 30, 40 percent) may be indicative of the desired value. The desired value may be set before the substantial change, proximate to the substantial change, and the like.
For example, referring to
Method 8220 is for managing a concept structure.
Method 8220 may include:
-
- Step 8222 of determining to add a new signature to the concept structure. The concept structure may already include at least one old signature. The new signature includes identifiers that identify at least parts of objects.
- Step 8224 of determining a new signature matching criterion that is based on one or more of the identifiers of the new signature. The new signature matching criterion determines when another signature matches the new signature. The determining of the new signature matching criterion may include evaluating object detection capabilities of the signature under different signature matching criteria.
Step 8224 may include steps 8203, 8204 and 8206 (include din step 8206) of method 8200.
EXAMPLES OF SYSTEMSThe system include various components, elements and/or units.
A component element and/or unit may be a processing circuitry may be implemented as a central processing unit (CPU), and/or one or more other integrated circuits such as application-specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), full-custom integrated circuits, etc., or a combination of such integrated circuits.
Alternatively, each component element and/or unit may implemented in hardware, firmware, or software that may be executed by a processing circuitry.
System 4900 may include sensing unit 4902, communication unit 4904, input 4911, processor 4950, and output 4919. The communication unit 4904 may include the input and/or the output.
Input and/or output may be any suitable communications component such as a network interface card, universal serial bus (USB) port, disk reader, modem or transceiver that may be operative to use protocols such as are known in the art to communicate either directly, or indirectly, with other elements of the system.
Processor 4950 may include at least some out of
-
- Multiple spanning elements 4951(q).
- Multiple merge elements 4952(r).
- Object detector 4953.
- Cluster manager 4954.
- Controller 4955.
- Selection unit 4956.
- Object detection determination unit 4957.
- Signature generator 4958.
- Movement information unit 4959.
- Identifier unit 4960.
There may be provided a method for low-power calculation of a signature, the method may include receiving or generating a media unit of multiple objects; processing the media unit by performing multiple iterations, wherein at least some of the multiple iterations may include applying, by spanning elements of the iteration, dimension expansion process that may be followed by a merge operation; wherein the applying of the dimension expansion process of an iteration may include determining a relevancy of the spanning elements of the iteration; completing the dimension expansion process by relevant spanning elements of the iteration and reducing a power consumption of irrelevant spanning elements until, at least, a completion of the applying of the dimension expansion process; determining identifiers that may be associated with significant portions of an output of the multiple iterations; and providing a signature that may include the identifiers and represents the multiple objects.
The identifiers may be retrieval information for retrieving the significant portions.
The at least some of the multiple iterations may be a majority of the multiple iterations.
The output of the multiple iteration may include multiple property attributes for each segment out of multiple segments of the media unit; and wherein the significant portions of an output of the multiple iterations may include more impactful property attributes.
A first iteration of the multiple iteration may include applying the dimension expansion process by applying different filters on the media unit.
The at least some of the multiple iteration exclude at least a first iteration of the multiple iterations.
The determining the relevancy of the spanning elements of the iteration may be based on at least some identities of relevant spanning elements of at least one previous iteration.
The determining the relevancy of the spanning elements of the iteration may be based on at least some identities of relevant spanning elements of at least one previous iteration that preceded the iteration.
The determining the relevancy of the spanning elements of the iteration may be based on properties of the media unit.
The determining the relevancy of the spanning elements of the iteration may be performed by the spanning elements of the iteration.
The method may include a neural network processing operation that may be executed by one or more layers of a neural network and does not belong to the at least some of the multiple iterations.
The at least one iteration may be executed without reducing power consumption of irrelevant neurons of the one or more layers.
The one or more layers output information about properties of the media unit, wherein the information differs from a recognition of the multiple objects.
The applying, by spanning elements of an iteration that differs from the first iteration, the dimension expansion process may include assigning output values that may be indicative of an identity of the relevant spanning elements of the iteration.
The applying, by spanning elements of an iteration that differs from the first iteration, the dimension expansion process may include assigning output values that may be indicative a history of dimension expansion processes until the iteration that differs from the first iteration.
Each spanning element may be associated with a subset of reference identifiers; and wherein the determining of the relevancy of each spanning elements of the iteration may be based a relationship between the subset of the reference identifiers of the spanning element and an output of a last merge operation before the iteration.
An output of a dimension expansion process of an iteration may be a multidimensional representation of the media unit that may include media unit regions of interest that may be associated with one or more expansion processes that generated the regions of interest.
A merge operation of the iteration may include selecting a subgroup of media unit regions of interest based on a spatial relationship between the subgroup of multidimensional regions of interest.
The method may include applying a merge function on the subgroup of multi dimensional regions of interest.
The method may include applying an intersection function on the subgroup of multidimensional regions of interest.
A merge operation of the iteration may be based on an actual size of one or more multidimensional regions of interest.
A merge operation of the iteration may be based on relationship between sizes of the multidimensional regions of interest.
A merge operation of the iteration may be based on changes of the media unit regions of interest during at least the iteration and one or more previous iteration.
There may be provided a non-transitory computer readable medium for low-power calculation of a signature, the non-transitory computer readable medium may store instructions for receiving or generating a media unit of multiple objects; processing the media unit by performing multiple iterations, wherein at least some of the multiple iterations may include applying, by spanning elements of the iteration, dimension expansion process that may be followed by a merge operation; wherein the applying of the dimension expansion process of an iteration may include determining a relevancy of the spanning elements of the iteration; completing the dimension expansion process by relevant spanning elements of the iteration and reducing a power consumption of irrelevant spanning elements until, at least, a completion of the applying of the dimension expansion process; determining identifiers that may be associated with significant portions of an output of the multiple iterations; and providing a signature that may include the identifiers and represents the multiple objects.
The identifiers may be retrieval information for retrieving the significant portions.
The at least some of the multiple iterations may be a majority of the multiple iterations.
The output of the multiple iteration may include multiple property attributes for each segment out of multiple segments of the media unit; and wherein the significant portions of an output of the multiple iterations may include more impactful property attributes.
A first iteration of the multiple iteration may include applying the dimension expansion process by applying different filters on the media unit.
The at least some of the multiple iteration exclude at least a first iteration of the multiple iterations.
The determining the relevancy of the spanning elements of the iteration may be based on at least some identities of relevant spanning elements of at least one previous iteration.
The determining the relevancy of the spanning elements of the iteration may be based on at least some identities of relevant spanning elements of at least one previous iteration that preceded the iteration.
The determining the relevancy of the spanning elements of the iteration may be based on properties of the media unit.
The determining the relevancy of the spanning elements of the iteration may be performed by the spanning elements of the iteration.
The non-transitory computer readable medium may store instructions for performing a neural network processing operation, by one or more layers of a neural network, wherein the neural network processing operation and does not belong to the at least some of the multiple iterations.
The non-transitory computer readable medium the at least one iteration may be executed without reducing power consumption of irrelevant neurons of the one or more layers.
The one or more layers output information about properties of the media unit, wherein the information differs from a recognition of the multiple objects.
The applying, by spanning elements of an iteration that differs from the first iteration, the dimension expansion process may include assigning output values that may be indicative of an identity of the relevant spanning elements of the iteration.
The applying, by spanning elements of an iteration that differs from the first iteration, the dimension expansion process may include assigning output values that may be indicative a history of dimension expansion processes until the iteration that differs from the first iteration.
Each spanning element may be associated with a subset of reference identifiers; and wherein the determining of the relevancy of each spanning elements of the iteration may be based a relationship between the subset of the reference identifiers of the spanning element and an output of a last merge operation before the iteration.
An output of a dimension expansion process of an iteration may be a multidimensional representation of the media unit that may include media unit regions of interest that may be associated with one or more expansion processes that generated the regions of interest.
A merge operation of the iteration may include selecting a subgroup of media unit regions of interest based on a spatial relationship between the subgroup of multidimensional regions of interest.
The non-transitory computer readable medium may store instructions for applying a merge function on the subgroup of multidimensional regions of interest.
The non-transitory computer readable medium may store instructions for applying an intersection function on the subgroup of multidimensional regions of interest.
A merge operation of the iteration may be based on an actual size of one or more multidimensional regions of interest.
A merge operation of the iteration may be based on relationship between sizes of the multidimensional regions of interest.
A merge operation of the iteration may be based on changes of the media unit regions of interest during at least the iteration and one or more previous iteration.
There may be provided a signature generator that may include an input that may be configured to receive or generate a media unit of multiple objects; an processor that may be configured to process the media unit by performing multiple iterations, wherein at least some of the multiple iterations may include applying, by spanning elements of the iteration, dimension expansion process that may be followed by a merge operation; wherein the applying of the dimension expansion process of an iteration may include determining a relevancy of the spanning elements of the iteration; completing the dimension expansion process by relevant spanning elements of the iteration and reducing a power consumption of irrelevant spanning elements until, at least, a completion of the applying of the dimension expansion process; an identifier unit that may be configured to determine identifiers that may be associated with significant portions of an output of the multiple iterations; and an output that may be configured to provide a signature that may include the identifiers and represents the multiple objects.
The identifiers may be retrieval information for retrieving the significant portions.
The at least some of the multiple iterations may be a majority of the multiple iterations.
The output of the multiple iteration may include multiple property attributes for each segment out of multiple segments of the media unit; and wherein the significant portions of an output of the multiple iterations may include more impactful property attributes.
A first iteration of the multiple iteration may include applying the dimension expansion process by applying different filters on the media unit.
The at least some of the multiple iteration exclude at least a first iteration of the multiple iterations.
The determining the relevancy of the spanning elements of the iteration may be based on at least some identities of relevant spanning elements of at least one previous iteration.
The determining the relevancy of the spanning elements of the iteration may be based on at least some identities of relevant spanning elements of at least one previous iteration that preceded the iteration.
The determining the relevancy of the spanning elements of the iteration may be based on properties of the media unit.
The determining the relevancy of the spanning elements of the iteration may be performed by the spanning elements of the iteration.
The signature generator that may include one or more layers of a neural network that may be configured to perform a neural network processing operation, wherein the neural network processing operation does not belong to the at least some of the multiple iterations.
At least one iteration may be executed without reducing power consumption of irrelevant neurons of the one or more layers.
The one or more layers output information about properties of the media unit, wherein the information differs from a recognition of the multiple objects.
The applying, by spanning elements of an iteration that differs from the first iteration, the dimension expansion process may include assigning output values that may be indicative of an identity of the relevant spanning elements of the iteration.
The applying, by spanning elements of an iteration that differs from the first iteration, the dimension expansion process may include assigning output values that may be indicative a history of dimension expansion processes until the iteration that differs from the first iteration.
Each spanning element may be associated with a subset of reference identifiers; and wherein the determining of the relevancy of each spanning elements of the iteration may be based a relationship between the subset of the reference identifiers of the spanning element and an output of a last merge operation before the iteration.
An output of a dimension expansion process of an iteration may be a multidimensional representation of the media unit that may include media unit regions of interest that may be associated with one or more expansion processes that generated the regions of interest.
A merge operation of the iteration may include selecting a subgroup of media unit regions of interest based on a spatial relationship between the subgroup of multidimensional regions of interest.
The signature generator that may be configured to apply a merge function on the subgroup of multidimensional regions of interest.
The signature generator that may be configured to apply an intersection function on the subgroup of multidimensional regions of interest.
A merge operation of the iteration may be based on an actual size of one or more multidimensional regions of interest.
A merge operation of the iteration may be based on relationship between sizes of the multidimensional regions of interest.
A merge operation of the iteration may be based on changes of the media unit regions of interest during at least the iteration and one or more previous iteration.
There may be provided There may be provided a method for low-power calculation of a signature of a media unit by a group of calculating elements, the method may include calculating, multiple attributes of segments of the media unit; wherein the calculating may include determining, by each calculating element of multiple calculating elements, a relevancy of the calculation unit to the media unit to provide irrelevant calculating elements and relevant calculating elements; reducing a power consumption of each irrelevant calculating element; and completing the calculating of the multiple attributes of segments of the media unit by the relevant calculating elements; determining identifiers that may be associated with significant attributes out of the multiple attributes of segments of the media unit; and providing a signature that may include the identifiers and represents the multiple objects.
The calculating elements may be spanning elements.
Each calculating element may be associated with a subset of one or more reference identifiers; and wherein the determining of a relevancy of the calculation unit to the media unit may be based on a relationship between the subset and the identifiers related to the media unit.
Each calculating element may be associated with a subset of one or more reference identifiers; and wherein a calculation element may be relevant to the media unit when the identifiers related to the media unit may include each reference identifier of the subset.
The calculating of the multiple attributes of segments of the media unit may be executed in multiple iterations; and wherein each iteration may be executed by calculation elements associated with the iteration; wherein the determining, by each calculating element of multiple calculating elements, of the relevancy of the calculation unit to the media unit may be executed per iteration.
The multiple iterations may be preceded by a calculation of initial media unit attributes by one or more layers of a neural network.
There may be provided a non-transitory computer readable medium for low-power calculation of a signature of a media unit by a group of calculating elements, the non-transitory computer readable medium may store instructions for calculating, multiple attributes of segments of the media unit; wherein the calculating may include determining, by each calculating element of multiple calculating elements, a relevancy of the calculation unit to the media unit to provide irrelevant calculating elements and relevant calculating elements; reducing a power consumption of each irrelevant calculating element; and completing the calculating of the multiple attributes of segments of the media unit by the relevant calculating elements; determining identifiers that may be associated with significant attributes out of the multiple attributes of segments of the media unit; and providing a signature that may include the identifiers and represents the multiple objects.
The calculating elements may be spanning elements.
Each calculating element may be associated with a subset of one or more reference identifiers; and wherein the determining of a relevancy of the calculation unit to the media unit may be based on a relationship between the subset and the identifiers related to the media unit.
Each calculating element may be associated with a subset of one or more reference identifiers; and wherein a calculation element may be relevant to the media unit when the identifiers related to the media unit may include each reference identifier of the subset.
The calculating of the multiple attributes of segments of the media unit may be executed in multiple iterations; and wherein each iteration may be executed by calculation elements associated with the iteration; wherein the determining, by each calculating element of multiple calculating elements, of the relevancy of the calculation unit to the media unit may be executed per iteration.
The multiple iterations may be preceded by a calculation of initial media unit attributes by one or more layers of a neural network.
There may be provided a signature generator that may include a processor that may be configured to calculate multiple attributes of segments of the media unit; wherein the calculating may include determining, by each calculating element of multiple calculating elements of the processor, a relevancy of the calculation unit to the media unit to provide irrelevant calculating elements and relevant calculating elements; reducing a power consumption of each irrelevant calculating element; and completing the calculating of the multiple attributes of segments of the media unit by the relevant calculating elements; an identifier unit that may be configured to determine identifiers that may be associated with significant attributes out of the multiple attributes of segments of the media unit; and an output that may be configured to provide a signature that may include the identifiers and represents the multiple objects.
The calculating elements may be spanning elements.
Each calculating element may be associated with a subset of one or more reference identifiers; and wherein the determining of a relevancy of the calculation unit to the media unit may be based on a relationship between the subset and the identifiers related to the media unit.
Each calculating element may be associated with a subset of one or more reference identifiers; and wherein a calculation element may be relevant to the media unit when the identifiers related to the media unit may include each reference identifier of the subset.
The calculating of the multiple attributes of segments of the media unit may be executed in multiple iterations; and wherein each iteration may be executed by calculation elements associated with the iteration; wherein the determining, by each calculating element of multiple calculating elements, of the relevancy of the calculation unit to the media unit may be executed per iteration.
The multiple iterations may be preceded by a calculation of initial media unit attributes by one or more layers of a neural network.
There may be provided a method for generating a hybrid representation of a media unit, the method may include receiving or generating the media unit; processing the media unit by performing multiple iterations, wherein at least some of the multiple iterations may include applying, by spanning elements of the iteration, dimension expansion process that may be followed by a merge operation; selecting, based on an output of the multiple iterations, media unit regions of interest that contributed to the output of the multiple iterations; and providing the hybrid representation, wherein the hybrid representation may include shape information regarding shapes of the media unit regions of interest, and a media unit signature that may include identifiers that identify the media unit regions of interest.
The selecting of the media regions of interest may be executed per segment out of multiple segments of the media unit.
The shape information may include polygons that represent shapes that substantially bound the media unit regions of interest.
The providing of the hybrid representation of the media unit may include compressing the shape information of the media unit to provide compressed shape information of the media unit.
The method may include comparing the media unit signature of the media unit to signatures of multiple concept structures to find a matching concept structure that has at least one matching signature that matches to the media unit signature; and calculating higher accuracy shape information that may be related to regions of interest of the media unit, wherein the higher accuracy shape information may be of higher accuracy than the compressed shape information of the media unit, wherein the calculating may be based on shape information associated with at least some of the matching signatures.
The method may include determining shapes of the media unit regions of interest using the higher accuracy shape information.
For each media unit region of interest, the calculating of the higher accuracy shape information may include virtually overlaying shapes of corresponding media units of interest of at least some of the matching signatures.
There may be provided a non-transitory computer readable medium for generating a hybrid representation of a media unit, the non-transitory computer readable medium may store instructions for receiving or generating the media unit; processing the media unit by performing multiple iterations, wherein at least some of the multiple iterations may include applying, by spanning elements of the iteration, dimension expansion process that may be followed by a merge operation; selecting, based on an output of the multiple iterations, media unit regions of interest that contributed to the output of the multiple iterations; and providing the hybrid representation, wherein the hybrid representation may include shape information regarding shapes of the media unit regions of interest, and a media unit signature that may include identifiers that identify the media unit regions of interest.
The selecting of the media regions of interest may be executed per segment out of multiple segments of the media unit.
The shape information may include polygons that represent shapes that substantially bound the media unit regions of interest.
The providing of the hybrid representation of the media unit may include compressing the shape information of the media unit to provide compressed shape information of the media unit.
The non-transitory computer readable medium may store instructions for comparing the media unit signature of the media unit to signatures of multiple concept structures to find a matching concept structure that has at least one matching signature that matches to the media unit signature; and calculating higher accuracy shape information that may be related to regions of interest of the media unit, wherein the higher accuracy shape information may be of higher accuracy than the compressed shape information of the media unit, wherein the calculating may be based on shape information associated with at least some of the matching signatures.
The non-transitory computer readable medium may store instructions for determining shapes of the media unit regions of interest using the higher accuracy shape information.
The for each media unit region of interest, the calculating of the higher accuracy shape information may include virtually overlaying shapes of corresponding media units of interest of at least some of the matching signatures.
There may be provided a hybrid representation generator for generating a hybrid representation of a media unit, the hybrid representation generator may include an input that may be configured to receive or generate the media unit; a processor that may be configured to process the media unit by by performing multiple iterations, wherein at least some of the multiple iterations may include applying, by spanning elements of the iteration, dimension expansion process that may be followed by a merge operation; a selection unit that may be configured to select, based on an output of the multiple iterations, media unit regions of interest that contributed to the output of the multiple iterations; and an output that may be configured to provide the hybrid representation, wherein the hybrid representation may include shape information regarding shapes of the media unit regions of interest, and a media unit signature that may include identifiers that identify the media unit regions of interest.
The selecting of the media regions of interest may be executed per segment out of multiple segments of the media unit.
The shape information may include polygons that represent shapes that substantially bound the media unit regions of interest.
The hybrid representation generator that may be configured to compress the shape information of the media unit to provide compressed shape information of the media unit.
There may be provided a method for scale invariant object detection, the method may include receiving or generating a first image in which an object appears in a first scale and a second image in which the object appears in a second scale that differs from the first scale; generating a first image signature and a second image signature; wherein the first image signature may include a first group of at least one certain first image identifier that identifies at least a part of the object; wherein the second image signature may include a second group of certain second image identifiers that identify different parts of the object; wherein the second group may be larger than first group; and linking between the at least one certain first image identifier and the certain second image identifiers.
The method may include linking between the first image signature, the second image signature and the object.
The linking may include adding the first signature and the second signature to a certain concept structure that may be associated with the object.
The method may include receiving or generating an input image; generating a signature of the input image; comparing the signature of the input image to signatures of the certain concept structure; and determining that the input image may include the object when at least one of the signatures of the certain concept structure matches the signature of the input image.
The method may include receiving or generating an input image; generating a signature of the input image, the signature of the input image may include only some of the certain second image identifiers; wherein the input image of the second scale; changing a scale of the input image to the first scale to a provide an amended input image; generating a signature of the amended input image; and verifying that the input image may include the object when the signature of the amended input image may include the at least one certain first image identifier.
The method may include receiving or generating an input image; generating a signature of the input image; searching in the signature of the input image for at least one of (a) the at least one certain first image identifier, and (b) the certain second image identifiers; and determining that the input image may include the object when the signature of the input image may include the at least one of (a) the at least one certain first image identifier, and (b) the certain second image identifiers.
The method may include receiving or generating an input image; generating a signature of the input image; searching in the signature of the input image for at least one of (a) one or more certain first image identifier of the at least one certain first image identifier, and (b) at least one certain second image identifier of the certain second image identifiers; and determining that a input image includes the object when the signature of the input image may include the at least one of (a) one or more certain first image identifier of the at least one certain first image identifier, and (b) the at least one certain second image identifier.
There may be provided a non-transitory computer readable medium for scale invariant object detection, the non-transitory computer readable medium may store instructions for receiving or generating a first image in which an object appears in a first scale and a second image in which the object appears in a second scale that differs from the first scale; generating a first image signature and a second image signature; wherein the first image signature may include a first group of at least one certain first image identifier that identifies at least a part of the object; wherein the second image signature may include a second group of certain second image identifiers that identify different parts of the object; wherein the second group may be larger than first group; and linking between the at least one certain first image identifier and the certain second image identifiers.
The non-transitory computer readable medium may store instructions for linking between the first image signature, the second image signature and the object.
The linking may include adding the first signature and the second signature to a certain concept structure that may be associated with the object.
The non-transitory computer readable medium may store instructions for receiving or generating an input image; generating a signature of the input image; comparing the signature of the input image to signatures of the certain concept structure; and determining that the input image may include the object when at least one of the signatures of the certain concept structure matches the signature of the input image.
The non-transitory computer readable medium may store instructions for receiving or generating an input image; generating a signature of the input image, the signature of the input image may include only some of the certain second image identifiers; wherein the input image of the second scale; changing a scale of the input image to the first scale to a provide an amended input image; generating a signature of the amended input image; and verifying that the input image may include the object when the signature of the amended input image may include the at least one certain first image identifier.
The non-transitory computer readable medium may store instructions for receiving or generating an input image; generating a signature of the input image; searching in the signature of the input image for at least one of (a) the at least one certain first image identifier, and (b) the certain second image identifiers; and determining that a input image includes the object when the signature of the input image may include the at least one of (a) the at least one certain first image identifier, and (b) the certain second image identifiers.
The non-transitory computer readable medium may store instructions for receiving or generating an input image; generating a signature of the input image; searching in the signature of the input image for at least one of (a) one or more certain first image identifier of the at least one certain first image identifier, and (b) at least one certain second image identifier of the certain second image identifiers; and determining that a input image includes the object when the signature of the input image may include the at least one of (a) one or more certain first image identifier of the at least one certain first image identifier, and (b) the at least one certain second image identifier.
There may be provided an object detector for scale invariant object detection, that may include an input that may be configured to receive a first image in which an object appears in a first scale and a second image in which the object appears in a second scale that differs from the first scale; a signature generator that may be configured to generate a first image signature and a second image signature; wherein the first image signature may include a first group of at least one certain first image identifier that identifies at least a part of the object; wherein the second image signature may include a second group of certain second image identifiers that identify different parts of the object; wherein the second group may be larger than first group; and an object detection determination unit that may be configured to link between the at least one certain first image identifier and the certain second image identifiers.
The object detection determination unit may be configured to link between the first image signature, the second image signature and the object.
The object detection determination unit may be configured to add the first signature and the second signature to a certain concept structure that may be associated with the object.
The input may be configured to receive an input image; wherein the signal generator may be configured to generate a signature of the input image; wherein the object detection determination unit may be configured to compare the signature of the input image to signatures of the certain concept structure, and to determine that the input image may include the object when at least one of the signatures of the certain concept structure matches the signature of the input image.
The input may be configured to receive an input image; wherein the signature generator may be configured to generate a signature of the input image, the signature of the input image may include only some of the certain second image identifiers; wherein the input image of the second scale; wherein the input may be configured to receive an amended input image that was generated by changing a scale of the input image to the first scale; wherein the signature generator may be configured to generate a signature of the amended input image; and wherein the object detection determination unit may be configured to verify that the input image may include the object when the signature of the amended input image may include the at least one certain first image identifier.
The input may be configured to receive an input image; wherein the signature generator may be configured to generate a signature of the input image; wherein the object detection determination unit may be configured to search in the signature of the input image for at least one of (a) the at least one certain first image identifier, and (b) the certain second image identifiers; and determine that a input image includes the object when the signature of the input image may include the at least one of (a) the at least one certain first image identifier, and (b) the certain second image identifiers.
The input may be configured to receive an input image; wherein the signature generator may be configured to generate a signature of the input image; wherein the object detection determination unit may be configured to search in the signature of the input image for at least one of (a) one or more certain first image identifier of the at least one certain first image identifier, and (b) at least one certain second image identifier of the certain second image identifiers; and determine that a input image includes the object when the signature of the input image may include the at least one of (a) one or more certain first image identifier of the at least one certain first image identifier, and (b) the at least one certain second image identifier.
There may be provided a method for movement based object detection, the method may include receiving or generating a video stream that may include a sequence of images; generating image signatures of the images; wherein each image may be associated with an image signature that may include identifiers; wherein each identifier identifiers a region of interest within the image; generating movement information indicative of movements of the regions of interest within consecutive images of the sequence of images; searching, based on the movement information, for a first group of regions of interest that follow a first movement; wherein different first regions of interest may be associated with different parts of an object; and linking between first identifiers that identify the first group of regions of interest.
The linking may include linking between first image signatures that include the first linked identifiers.
The linking may include adding the first image signatures to a first concept structure, the first concept structure may be associated with the first image.
The method may include receiving or generating an input image; generating a signature of the input image; comparing the signature of the input image to signatures of the first concept structure; and determining that the input image may include a first object when at least one of the signatures of the first concept structure matches the signature of the input image.
The method may include receiving or generating an input image; generating a signature of the input image; searching in the signature of the input image for at least one of the first identifiers; and determining that the input image may include the object when the signature of the input image may include at least one of the first identifiers.
The method may include generating location information indicative of a location of each region of interest within each image; wherein the generating movement information may be based on the location information.
There may be provided a non-transitory computer readable medium for movement based object detection, the non-transitory computer readable medium may include receiving or generating a video stream that may include a sequence of images; generating image signatures of the images; wherein each image may be associated with an image signature that may include identifiers; wherein each identifier identifiers a region of interest within the image; wherein different region of interests include different objects; generating movement information indicative of movements of the regions of interest within consecutive images of the sequence of images; searching, based on the movement information, for a first group of regions of interest that follow a first movement; and linking between first identifiers that identify the first group of regions of interest.
The linking may include linking between first image signatures that include the first linked identifiers.
The linking may include adding the first image signatures to a first concept structure, the first concept structure may be associated with the first image.
The non-transitory computer readable medium may store instructions for receiving or generating an input image; generating a signature of the input image; comparing the signature of the input image to signatures of the first concept structure; and determining that the input image may include a first object when at least one of the signatures of the first concept structure matches the signature of the input image.
The non-transitory computer readable medium may store instructions for receiving or generating an input image; generating a signature of the input image; searching in the signature of the input image for at least one of the first identifiers; and determining that the input image may include the object when the signature of the input image may include at least one of the first identifiers.
The non-transitory computer readable medium may store instructions for generating location information indicative of a location of each region of interest within each image; wherein the generating movement information may be based on the location information.
There may be provided an object detector that may include an input that may be configured to receive a video stream that may include a sequence of images; a signature generator that may be configured to generate image signatures of the images; wherein each image may be associated with an image signature that may include identifiers; wherein each identifier identifiers a region of interest within the image; a movement information unit that may be is configured to generate movement information indicative of movements of the regions of interest within consecutive images of the sequence of images; an object detection determination unit that may be configured to search, based on the movement information, for a first group of regions of interest that follow a first movement; wherein different first regions of interest may be associated with different parts of an object; and link between first identifiers that identify the first group of regions of interest.
The linking may include linking between first image signatures that include the first linked identifiers.
The linking may include adding the first image signatures to a first concept structure, the first concept structure may be associated with the first image.
The input may be configured to receive an input image; wherein the signature generator may be configured to generate a signature of the input image; and wherein the object detection determination unit may be configured to compare the signature of the input image to signatures of the first concept structure; and to determine that the input image may include a first object when at least one of the signatures of the first concept structure matches the signature of the input image.
The input may be configured to receive an input image; wherein the signature generator may be configured to generate a signature of the input image; and wherein the object detection determination unit may be configured to search in the signature of the input image for at least one of the first identifiers; and to determine that the input image may include the object when the signature of the input image may include at least one of the first identifiers.
The object detector that may be configured to generate location information indicative of a location of each region of interest within each image; wherein the generating of the movement information may be based on the location information.
There may be provided a method for object detection, the method may include receiving an input image; generating a signature of the input image; comparing the signature of the input image to signatures of a concept structure; determining whether the signature of the input image matches any of the signatures of the concept structure based on signature matching criteria, wherein each signature of the concept structure may be associated within a signature matching criterion that may be determined based on an object detection parameter of the signature; and concluding that the input image may include an object associated with the concept structure based on an outcome of the determining.
Each signature matching criterion may be determined by evaluating object detection capabilities of the signature under different signature matching criteria.
The evaluating of the object detection capabilities of the signature under different signature matching criteria may include receiving or generating signatures of a group of test images; calculating the object detection capability of the signature, for each signature matching criterion of the different signature matching criteria; and selecting the signature matching criterion based on the object detection capabilities of the signature under the different signature matching criteria.
The object detection capability reflects a percent of signatures of the group of test images that match the signature.
The selecting of the signature matching criterion may include selecting the signature matching criterion that one applied results in a percent of signatures of the group of test images that match the signature that may be closets to a predefined desired percent of signatures of the group of test images that match the signature.
The signature matching criteria may be a minimal number of matching identifiers that indicate of a match.
There may be provided a method for managing a concept structure, the method may include determining to add a new signature to the concept structure, wherein the concept structure already may include at least one old signature; wherein the new signature may include identifiers that identify at least parts of objects; and determining a new signature matching criterion that may be based on one or more of the identifiers of the new signature; wherein the new signature matching criterion determines when another signature matches the new signature; wherein the determining of the new signature matching criterion may include evaluating object detection capabilities of the signature under different signature matching criteria.
The evaluating of the object detection capabilities of the signature under different signature matching criteria may include receiving or generating signatures of a group of test images; calculating the object detection capability of the signature, for each signature matching criterion of the different signature matching criteria; and selecting the signature matching criterion based on the object detection capabilities of the signature under the different signature matching criteria.
The object detection capability reflects a percent of signatures of the group of test images that match the signature.
The selecting of the signature matching criterion may include selecting the signature matching criterion that one applied results in a percent of signatures of the group of test images that match the signature that may be closets to a predefined desired percent of signatures of the group of test images that match the signature.
The signature matching criteria may be a minimal number of matching identifiers that indicate of a match.
There may be provided a non-transitory computer readable medium for object detection, the non-transitory computer readable medium may store instructions for receiving an input image; generating a signature of the input image; comparing the signature of the input image to signatures of a concept structure; determining whether the signature of the input image matches any of the signatures of the concept structure based on signature matching criteria, wherein each signature of the concept structure may be associated within a signature matching criterion that may be determined based on an object detection parameter of the signature; and concluding that the input image may include an object associated with the concept structure based on an outcome of the determining.
Each signature matching criterion may be determined by evaluating object detection capabilities of the signature under different signature matching criteria.
The evaluating of the object detection capabilities of the signature under different signature matching criteria may include receiving or generating signatures of a group of test images; calculating the object detection capability of the signature, for each signature matching criterion of the different signature matching criteria; and selecting the signature matching criterion based on the object detection capabilities of the signature under the different signature matching criteria.
The object detection capability reflects a percent of signatures of the group of test images that match the signature.
The selecting of the signature matching criterion may include selecting the signature matching criterion that one applied results in a percent of signatures of the group of test images that match the signature that may be closets to a predefined desired percent of signatures of the group of test images that match the signature.
The signature matching criteria may be a minimal number of matching identifiers that indicate of a match.
There may be provided a non-transitory computer readable medium for managing a concept structure, the non-transitory computer readable medium may store instructions for determining to add a new signature to the concept structure, wherein the concept structure already may include at least one old signature; wherein the new signature may include identifiers that identify at least parts of objects; and determining a new signature matching criterion that may be based on one or more of the identifiers of the new signature; wherein the new signature matching criterion determines when another signature matches the new signature; wherein the determining of the new signature matching criterion may include evaluating object detection capabilities of the signature under different signature matching criteria.
The evaluating of the object detection capabilities of the signature under different signature matching criteria may include receiving or generating signatures of a group of test images; calculating the object detection capability of the signature, for each signature matching criterion of the different signature matching criteria; and selecting the signature matching criterion based on the object detection capabilities of the signature under the different signature matching criteria.
The object detection capability reflects a percent of signatures of the group of test images that match the signature.
The selecting of the signature matching criterion may include selecting the signature matching criterion that one applied results in a percent of signatures of the group of test images that match the signature that may be closets to a predefined desired percent of signatures of the group of test images that match the signature.
The signature matching criteria may be a minimal number of matching identifiers that indicate of a match.
There may be provided an object detector that may include an input that may be configured to receive an input image; a signature generator that may be configured to generate a signature of the input image; an object detection determination unit that may be configured to compare the signature of the input image to signatures of a concept structure; determine whether the signature of the input image matches any of the signatures of the concept structure based on signature matching criteria, wherein each signature of the concept structure may be associated within a signature matching criterion that may be determined based on an object detection parameter of the signature; and conclude that the input image may include an object associated with the concept structure based on an outcome of the determining.
The object detector according to claim, may include a signature matching criterion unit that may be configured to determine each signature matching criterion by evaluating object detection capabilities of the signature under different signature matching criteria.
The input may be configured to receive signatures of a group of test images; wherein the signature matching criterion unit may be configured to calculate the object detection capability of the signature, for each signature matching criterion of the different signature matching criteria; and select the signature matching criterion based on the object detection capabilities of the signature under the different signature matching criteria.
The object detector according to claim wherein the object detection capability reflects a percent of signatures of the group of test images that match the signature.
The object detector according to claim wherein the signature matching criterion unit may be configured to select the signature matching criterion that one applied results in a percent of signatures of the group of test images that match the signature that may be closets to a predefined desired percent of signatures of the group of test images that match the signature.
The object detector according to claim wherein the signature matching criteria may be a minimal number of matching identifiers that indicate of a match.
There may be provided a concept structure manager that may include a controller that may be configured to determine to add a new signature to the concept structure, wherein the concept structure already may include at least one old signature; wherein the new signature may include identifiers that identify at least parts of objects; and a signature matching criterion unit that may be configured to determine a new signature matching criterion that may be based on one or more of the identifiers of the new signature; wherein the new signature matching criterion determines when another signature matches the new signature; wherein the determining of the new signature matching criterion may include evaluating object detection capabilities of the signature under different signature matching criteria.
The signature matching criterion unit may be configured to determine each signature matching criterion by evaluating object detection capabilities of the signature under different signature matching criteria.
The input may be configured to receive signatures of a group of test images; wherein the signature matching criterion unit may be configured to calculate the object detection capability of the signature, for each signature matching criterion of the different signature matching criteria; and select the signature matching criterion based on the object detection capabilities of the signature under the different signature matching criteria.
The concept manager according to claim wherein the object detection capability reflects a percent of signatures of the group of test images that match the signature.
The concept manager according to claim wherein the signature matching criterion unit may be configured to select the signature matching criterion that one applied results in a percent of signatures of the group of test images that match the signature that may be closets to a predefined desired percent of signatures of the group of test images that match the signature.
The concept manager according to claim wherein the signature matching criteria may be a minimal number of matching identifiers that indicate of a match.
There may be provided a system, method and a non-transitory computer readable medium for predicting a movement of a hybrid behavior vehicle.
The prediction may include a first phase of generating an environment-based behavioral model of a hybrid-behavior vehicle and a second phase of using the environment-based behavioral model for predicting a movement of a hybrid behavior vehicle.
Information gained during the second phase may be used to update the model generate during the first phase. The first phase may be repeated multiple times, may be triggered by various events, lapse of time, and the like. For example, errors in the prediction of one or more predicted behaviors of a hybrid-behavior vehicle may trigger an evaluation of the model.
The first and second phases may be executed by the same entity or by different entities. The entities may be a vehicle computer, a remote computerized system, and the like.
Method 100 may start by step 110 of receiving video streams of driving scenes. A driving scene is a scene that is related to driving. A driving scene may be captured by a vehicle during a driving session. A driving scene may be captured by an a sensor located outside a vehicle—but the driving scene should include one or more vehicles, may include a road or any other path through which one or more vehicles may progress. A driving scene may include an environment of a vehicle.
The video streams may include a vast amount of information—even may be regarded as big data. The vast amount of information may include video streams of an aggregate durations of days, weeks, months, years and the like. Usually the accuracy of the prediction may improve with the increment of the amount of video streams.
Step 110 may be followed by step 120 of analyzing the video streams by creating an environment-based behavioral model of each hybrid-behavior vehicle of the vehicles to provide multiple environment-based behavioral models.
Step 120 may include an unsupervised learning process applied by a machine learning process.
Step 120 may include calculating signatures to each received image, comparing the signatures to concept structures to identify objects, obtaining motion information regarding the objects, and the like.
Each environment-based behavioral model associates environmental elements located at an environment of a hybrid-behavior vehicle to one or more predicted behaviors of the hybrid behavior vehicle.
Each environment-based behavioral model may also take into account a current behavior of the hybrid behavior vehicle. For example—the environment-based behavioral model may differentiate between different current behaviors. For example—the current behavior of the hybrid behavior vehicle may affect the predicted path, the probability associated with any of the predicted paths and the like. The current behavior of the hybrid behavior vehicle may be regarded as one of the environmental parameters.
The model may also assign a prediction certainty (or other score) to each one of the one or more predicted behaviors. This may indicate the probability or the chances that the hybrid-behavior vehicle will actually follow each predicted behavior.
It should be noted that each environment-based behavioral model may be responsive to additional parameters such as a geographical position. For example—The drivers of hybrid-behavior vehicles at one city may be less aggressive than drivers of another city. Yet for another example—different models of the hybrid-behavior vehicles may be distributed at different locations (due to regulation, cost or any other reason). The environment-based behavioral models may include the additional information and may be generated in an unsupervised manner.
Each environment-based behavioral model associates environmental elements located at an environment of a hybrid-behavior vehicle to one or more predicted behaviors of the hybrid behavior vehicle
These environment-based behavioral models provide a more accurate prediction of the movement of a hybrid behavior vehicle. Furthermore—the environmental-based behavioral models may distill the relevant environmental elements that may actually determine the predicted behavior and as such may be more compact than less accurate parameters that store more memory resource and their processing is less efficient.
In order to reduce the consumption of memory and computational resources, step 120 may include step 122 of recognizing vehicles that appear in the video streams, step 123 of detecting non-hybrid-behavior vehicles, and step 124 of preventing form building behavioral models of non-hybrid-behavior vehicles.
Step 120 may include step 121 of classifying the hybrid-behavior vehicles to different classes classes of hybrid-behavior vehicles.
The classification may be followed by building a model for each class and/or by generating metadata that may augment a model that may be shared between different classes.
Using a model per class may provide more accurate prediction of the predicting a movement of a hybrid behavior vehicle.
The classification may be based on various parameters, such as but not limited to (a) visual attributes of the hybrid-behavior vehicles, (b) movement parameters of the hybrid-behavior vehicles, and (c) risk levels associated with movements of the hybrid-behavior vehicles.
The visual attributes may refer to any one of the shape, size, color, and or any other identifier (for example—the logo of the manufacturer, name of the model, and the like) of the hybrid-behavior vehicles. A unique shape and/or size and/or color can be easily detected.
The movement parameters may related to speed, acceleration, breaking distance, deceleration, radius of turn, and any movement patterns. The movement parameters may be determined by the physical characters of the hybrid-behavior vehicle (engine, chassis, gear) and to the drivers themselves.
The risk level may refer to the risk level associated with driving the hybrid-behavior vehicle. For example—hybrid-behavior vehicles that are more popular among younger drivers may exhibit more aggressive movement parameters—associated with higher risk—than those exhibited by seasoned drivers.
The classification and the environment-based behavioral models may be built on the fly—and/or based on sensed information regarding the behavior of the hybrid-behavior vehicles—and thus can easily learn new classes of hybrid-behavior vehicles—even without any prior knowledge about the hybrid-behavior vehicles.
Step 120 may be followed by step 130 of responding to the creation of the multiple environment-based behavioral models.
The responding may include storing the multiple environment-based behavioral models. Each environment-based behavioral model may be associated with a certain class of hybrid-behavior vehicle.
The responding may include sending the multiple environment-based behavioral models to vehicles or to any other entities.
Method 200 may start by step 210 of receiving an environment-based behavioral model of a hybrid-behavior vehicle.
Step 210 may include receiving multiple environment-based behavioral models of multiple hybrid-behavior vehicles.
The reception may be based on the location of the hybrid-behavior vehicle—as it will be more memory and cost effective to receive environment-based behavioral models that are relevant to the location of the autonomous vehicle.
Step 210 may be followed by step 220 of sensing an environment of the autonomous vehicle by one or more autonomous vehicle sensor.
Step 220 may be followed by step 230 of searching, in the environment of the autonomous vehicle, environmental elements and a hybrid-behavior vehicle.
If there are no environmental elements and no hybrid-behavior vehicle in the environment—jumping back to step 230.
If there is an hybrid-behavior vehicle in the environment and there are environmental elements then step 230 is followed by step 240 of predicting one or more predicted behaviors of the hybrid behavior vehicle based on the environmental elements and an environment-based behavioral model of the hybrid-behavior vehicle.
Step 240 may be followed by step 250 of driving the autonomous vehicle based, at least in part, on the one or more predicted behaviors of the hybrid behavior vehicle.
For example—driving the autonomous vehicle such as to avoid any contact with the hybrid-behavior vehicle.
Method 200 may start by step 210 of receiving an environment-based behavioral model of a hybrid-behavior vehicle.
Step 210 may include receiving multiple environment-based behavioral models of multiple hybrid-behavior vehicles.
The reception may be based on the location of the hybrid-behavior vehicle—as it will be more memory and cost effective to receive environment-based behavioral models that are relevant to the location of the autonomous vehicle.
Step 210 may be followed by step 220 of sensing an environment of the autonomous vehicle by one or more autonomous vehicle sensor.
Step 220 may be followed by step 230 of searching, in the environment of the autonomous vehicle, environmental elements and a hybrid-behavior vehicle.
If there are no environmental elements and no hybrid-behavior vehicle in the environment—jumping back to step 230.
If there is an hybrid-behavior vehicle in the environment and there are environmental elements then step 230 is followed by step 240 of predicting one or more predicted behaviors of the hybrid behavior vehicle based on the environmental elements and an environment-based behavioral model of the hybrid-behavior vehicle.
Step 240 may be followed by step 252 responding to the estimate.
Step 252 may include at least one out of
-
- Step 253 of providing to the driver an indication about the one or more predicted behaviors of the hybrid behavior vehicle.
- Step 254 of calculating a suggested propagation path of the vehicle based on the one or more predicted behaviors of the hybrid behavior vehicle.
- Step 255 of providing to the driver an indication about the suggested propagation path of the vehicle.
The driving scene also includes a first vehicle 414 that drives along the first lane 411 and a second vehicle 415 that drives along the second lane 412 and at an opposite direction to the certain vehicle.
The driving scene also includes objects such as trees, buildings, a cloud the pour rain on the road.
The driving scene also includes a person mounted on a hybrid-behavior vehicle such as a kick scooter 440.
Method 100 may include detecting the kick scooter, extracting at least one out of visual attributes of the kick scooter 440, movement parameters of the kick scooter 440, and risk level associated with movements of the kick scooter 440. Data obtained from multiple monitoring attempts of multiple kick scooters may be processed to provide an environment-based behavioral model of the kick scooter.
Each one of methods 200 and 202 may (a) receive an environment-based behavioral model of the kick scooter, and (b) based on environmental elements such as the road 410, building located near the road, the first and second vehicles and their spatial relationship to the kick scooter 440, provide one or more predicted behaviors of the hybrid behavior vehicle.
In
Each predicted behavior may be associated with a probability of occurrence.
It should be noted that each predicted behavior may including timing and/or velocity information indicative of the predicted speed (or speeds) of progress along the predicted path.
These suggested propagation paths may be suggested to a driver of a vehicle or may be applied by the autonomous vehicle.
Roundabout 520 that has three arms 511, 512 and 512. The roundabout is preceded by a roundabout related traffic sign 571.
The roundabout 520 includes an interior circle 521 and has two lanes. There are few pedestrians that cross the third arm 512 over a second zebra crossing 551 while a first zebra crossing 551 is located in front of the certain vehicle.
A person that drives a motorized kick scooters 440′ drives in the roundabout and is located to the left of the first arm 551.
The environmental elements that may be taken into account in the prediction may be the roundabout, the certain vehicle, the first and second zebra crossings, the first, second and third arms, the spatial relationship between the roundabout, the certain vehicle, the first and second zebra crossings, the first, second and third arms, and any one of the roundabout, the certain vehicle, the first and second zebra crossings, the first, second and third arms, and the like.
Method 100 may include detecting the motorized kick scooter 440′, extracting at least one out of visual attributes of the motorized kick scooter 440′, movement parameters of the motorized kick scooter 440′, and risk level associated with movements of the motorized kick scooter 440′. Data obtained from multiple monitoring attempts of multiple kick scooters may be processed to provide an environment-based behavioral model of the kick scooter.
Each one of methods 200 and 202 may (a) receive an environment-based behavioral model of the motorized kick scooter 440′, and (b) based on environmental elements such as the roundabout, the certain vehicle, the first and second zebra crossings, the first, second and third arms, provide one or more predicted behaviors of the hybrid behavior vehicle.
In
Each predicted behavior may be associated with a probability of occurrence.
It should be noted that each predicted behavior may including timing and/or velocity information indicative of the predicted speed (or speeds) of progress along the predicted path.
Reference is now made to
Processing circuitry 710 may be operative to execute instructions stored in memory (not shown). For example, processing circuitry 710 may be operative to execute method 200.
It will be appreciated that processing circuitry 710 may be implemented as a central processing unit (CPU), and/or one or more other integrated circuits such as application-specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), full-custom integrated circuits, etc., or a combination of such integrated circuits. It will similarly be appreciated that autonomous driving system 700 may comprise more than one instance of processing circuitry 710. For example, one such instance of processing circuitry 710 may be a special purpose processor operative to execute autonomous driving manager 760 to perform some, or all, of the functionality of autonomous driving system 700 as described herein.
I/O module 720 may be any suitable communications component such as a network interface card, universal serial bus (USB) port, disk reader, modem or transceiver that may be operative to use protocols such as are known in the art to communicate either directly, or indirectly, with other elements of system 700, such as, for example, remote computerized system (such as prediction server) 702, camera 730, telemetry ECU 740. As such, I/O module 720 may be operative to use a wired or wireless connection to connect to remote computerized system 702 via a communications network such as a local area network, a backbone network and/or the Internet, etc. I/O module 720 may also be operative to use a wired or wireless connection to connect to other components of system 700, e.g., camera 730, telemetry ECU 740. It will be appreciated that in operation I/O module 720 may be implemented as a multiplicity of modules, where different modules may be operative to use different communication technologies. For example, a module providing mobile network connectivity may be used to connect to remote computerized system 702, whereas a local area wired connection may be used to connect to camera 730, telemetry ECU 740.
In accordance with embodiments described herein, camera 730, telemetry ECU 740, and represent implementations of sensor(s) 130. It will be appreciated that camera 730, telemetry ECU 740, may be implemented as integrated components of vehicle 701 and may provide other functionality that is the interests of clarity is not explicitly described herein. As described hereinbelow, system 700 may use information about a current driving environment as received from camera 730, telemetry ECU 740, to determine an appropriate driving policy for vehicle 701.
Autonomous driving manager 760 may be an application implemented in hardware, firmware, or software that may be executed by processing circuitry 710 to provide driving instructions to vehicle 701. For example, autonomous driving manager 760 may use images received from camera 730 and/or telemetry data received from telemetry ECU 740 to determine an appropriate driving policy for arriving at a given destination and provide driving instructions to vehicle 701 accordingly. It will be appreciated that autonomous driving manager 760 may also be operative to use other data sources when determining a driving policy, e.g., maps of potential routes, traffic congestion reports, etc.
System 700 may also include a hybrid-behavior vehicle detector 762 for detecting, in the environment of the autonomous vehicle, environmental elements and a hybrid-behavior vehicle.
System 700 may also include a prediction module 764 that is configured to predict one or more predicted behaviors of the hybrid behavior vehicle based on the environmental elements and an environment-based behavioral model of the hybrid-behavior vehicle.
The autonomous driving manager 760 may be configured to predict one or more predicted behaviors of the hybrid behavior vehicle based on the environmental elements and an environment-based behavioral model of the hybrid-behavior vehicle; and an autonomous driving pattern unit 766 that is configured to determine a response of the autonomous vehicle to the one or more predicted behaviors of the hybrid behavior vehicle, and to control or at least trigger a driving of the autonomous vehicle based, at least in part, on the one or more predicted behaviors of the hybrid behavior vehicle.
System 700 also includes a database 770 for storing metadata such as multiple environment-based behavioral models of multiple hybrid-behavior vehicles.
It should be noted that that the vehicle may include, in addition to system 700 or instead of system 700 an advanced driver assistant system (702) that may differ from system 700 by assisting a driver in driving the vehicle in response to the one or more predicted behaviors of the hybrid behavior vehicle.
Reference is now made to
Each one of the modules may be instantiated in a suitable memory for storing software such as, for example, an optical storage medium, a magnetic storage medium, an electronic storage medium, and/or a combination thereof.
Processing circuitry 810 may be operative to execute instructions stored in memory (not shown). For example, processing circuitry 810 may be operative to execute code, instructions, application for executing method 100. It will be appreciated that processing circuitry 810 may be implemented as a central processing unit (CPU), and/or one or more other integrated circuits such as application-specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), full-custom integrated circuits, etc., or a combination of such integrated circuits. It will similarly be appreciated that server 702 may comprise more than one instance of processing circuitry 810. For example, one such instance of processing circuitry 810 may be a special purpose processor operative to execute any of the modules 830.
I/O module 820 may be any suitable communications component such as a network interface card, universal serial bus (USB) port, disk reader, modem or transceiver that may be operative to use protocols such as are known in the art to communicate either directly, or indirectly, with other elements of system 700 (or 702) (of
Each one of modules 830 may be hardware, firmware, or software.
Hybrid-behavior vehicle detector 832 may be configured to detecting non-hybrid-behavior vehicles in video streams.
Classifier 834 may be configured to classify the hybrid-behavior vehicles to different classes classes of hybrid-behavior vehicles.
Environment-based behavioral model generator 836 may be configured to create an environment-based behavioral model of each hybrid-behavior vehicle of the vehicles to provide multiple environment-based behavioral models.
Suggested response generator 838 is configured to determine a response—informing a driver, calculating responses to be suggested to a driver or to be implemented by an autonomous vehicle.
While the foregoing written description of the invention enables one of ordinary skill to make and use what is considered presently to be the best mode thereof, those of ordinary skill will understand and appreciate the existence of variations, combinations, and equivalents of the specific embodiment, method, and examples herein. The invention should therefore not be limited by the above described embodiment, method, and examples, but by all embodiments and methods within the scope and spirit of the invention as claimed.
In the foregoing specification, the invention has been described with reference to specific examples of embodiments of the invention. It will, however, be evident that various modifications and changes may be made therein without departing from the broader spirit and scope of the invention as set forth in the appended claims.
Moreover, the terms “front,” “back,” “top,” “bottom,” “over,” “under” and the like in the description and in the claims, if any, are used for descriptive purposes and not necessarily for describing permanent relative positions. It is understood that the terms so used are interchangeable under appropriate circumstances such that the embodiments of the invention described herein are, for example, capable of operation in other orientations than those illustrated or otherwise described herein.
Furthermore, the terms “assert” or “set” and “negate” (or “deassert” or “clear”) are used herein when referring to the rendering of a signal, status bit, or similar apparatus into its logically true or logically false state, respectively. If the logically true state is a logic level one, the logically false state is a logic level zero. And if the logically true state is a logic level zero, the logically false state is a logic level one.
Those skilled in the art will recognize that the boundaries between logic blocks are merely illustrative and that alternative embodiments may merge logic blocks or circuit elements or impose an alternate decomposition of functionality upon various logic blocks or circuit elements. Thus, it is to be understood that the architectures depicted herein are merely exemplary, and that in fact many other architectures may be implemented which achieve the same functionality.
Any arrangement of components to achieve the same functionality is effectively “associated” such that the desired functionality is achieved. Hence, any two components herein combined to achieve a particular functionality may be seen as “associated with” each other such that the desired functionality is achieved, irrespective of architectures or intermedial components. Likewise, any two components so associated can also be viewed as being “operably connected,” or “operably coupled,” to each other to achieve the desired functionality.
Furthermore, those skilled in the art will recognize that boundaries between the above described operations merely illustrative. The multiple operations may be combined into a single operation, a single operation may be distributed in additional operations and operations may be executed at least partially overlapping in time. Moreover, alternative embodiments may include multiple instances of a particular operation, and the order of operations may be altered in various other embodiments.
Also for example, in one embodiment, the illustrated examples may be implemented as circuitry located on a single integrated circuit or within a same device. Alternatively, the examples may be implemented as any number of separate integrated circuits or separate devices interconnected with each other in a suitable manner.
However, other modifications, variations and alternatives are also possible. The specifications and drawings are, accordingly, to be regarded in an illustrative rather than in a restrictive sense.
In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word ‘comprising’ does not exclude the presence of other elements or steps then those listed in a claim. Furthermore, the terms “a” or “an,” as used herein, are defined as one or more than one. Also, the use of introductory phrases such as “at least one” and “one or more” in the claims should not be construed to imply that the introduction of another claim element by the indefinite articles “a” or “an” limits any particular claim containing such introduced claim element to inventions containing only one such element, even when the same claim includes the introductory phrases “one or more” or “at least one” and indefinite articles such as “a” or “an.” The same holds true for the use of definite articles. Unless stated otherwise, terms such as “first” and “second” are used to arbitrarily distinguish between the elements such terms describe. Thus, these terms are not necessarily intended to indicate temporal or other prioritization of such elements. The mere fact that certain measures are recited in mutually different claims does not indicate that a combination of these measures cannot be used to advantage.
While certain features of the invention have been illustrated and described herein, many modifications, substitutions, changes, and equivalents will now occur to those of ordinary skill in the art. It is, therefore, to be understood that the appended claims are intended to cover all such modifications and changes as fall within the true spirit of the invention.
It is appreciated that various features of the embodiments of the disclosure which are, for clarity, described in the contexts of separate embodiments may also be provided in combination in a single embodiment. Conversely, various features of the embodiments of the disclosure which are, for brevity, described in the context of a single embodiment may also be provided separately or in any suitable sub-combination.
It will be appreciated by persons skilled in the art that the embodiments of the disclosure are not limited by what has been particularly shown and described hereinabove. Rather the scope of the embodiments of the disclosure is defined by the appended claims and equivalents thereof.
Claims
1. A method for predicting a movement of a hybrid-behavior vehicle, the method comprises:
- receiving, by a computerized system, video streams of driving scenes;
- analyzing, by the computerized system, the video streams; wherein the analyzing comprises: recognizing vehicles that appear in the video streams; wherein the vehicles comprise the hybrid-behavior vehicle and a non-hybrid-behavior vehicle; wherein the non-hybrid-behavior vehicle differs from the hybrid-behavior vehicle, wherein the hybrid-behavior vehicle is selected out of a group consisting of an unmotorized bicycle, a motorized bicycle, a motorized skateboard, a unicycle, a hoverboards and a kick scooter; detecting the non-hybrid-behavior vehicle and the hybrid-behavior vehicle; creating an environment-based behavioral model of the hybrid-behavior vehicle; preventing from building behavioral models of the non-hybrid-behavior vehicle; and
- responding, by the computerized system, to the creation of the environment-based behavioral model.
2. The method according to claim 1 comprising generating, by the computerized system, the environment-based behavioral model of the hybrid-behavior vehicle even without prior knowledge about the hybrid-behavior vehicle.
3. The method according to claim 1 wherein the vehicles that appear in the video stream comprise different classes of hybrid-behavior vehicles; wherein the method comprises creating different environment-based behavioral models for the different classes.
4. The method according to claim 3 wherein the analyzing comprises classifying the hybrid-behavior vehicles to the different classes based on visual attributes of the hybrid-behavior vehicles.
5. The method according to claim 3 wherein the analyzing comprises classifying the hybrid-behavior vehicles to the different classes based on movement parameters of the hybrid-behavior vehicles.
6. The method according to claim 3 wherein the analyzing comprises classifying the hybrid-behavior vehicles to the different classes based on risk levels associated with movements of the hybrid-behavior vehicles.
7. The method according to claim 3 wherein the analyzing comprises classifying the hybrid-behavior vehicles to the different classes based on at least two out of (a) visual attributes of the hybrid-behavior vehicles, (b) movement parameters of the hybrid-behavior vehicles, and (c) risk levels associated with movements of the hybrid-behavior vehicles.
8. A non-transitory computer readable medium for predicting a movement of a hybrid behavior vehicle, the non-transitory computer readable medium stores instructions for:
- receiving, by a computerized system, video streams of driving scenes;
- analyzing, by the computerized system, the video streams; wherein the analyzing comprises: recognizing vehicles that appear in the video streams; wherein the vehicles comprise the hybrid-behavior vehicle and a non-hybrid-behavior vehicle; wherein the non-hybrid-behavior vehicle differs from the hybrid-behavior vehicle, wherein the hybrid-behavior vehicle is selected out of a group consisting of an unmotorized bicycle, a motorized bicycle, a motorized skateboard, a unicycle, a hoverboards and a kick scooter; detecting the non-hybrid-behavior vehicle and the hybrid-behavior vehicle; creating an environment-based behavioral model of the hybrid-behavior vehicle; preventing from building behavioral models of the non-hybrid-behavior vehicle; and responding, by the computerized system, to the creation of the environment-based behavioral model.
9. The non-transitory computer readable medium according to claim 8 that stores instructions for generating the environment-based behavioral model of the hybrid behavior vehicle even without prior knowledge about the hybrid behavior vehicle.
10. The non-transitory computer readable medium according to claim 8 wherein the vehicles that appear in the video stream comprise different classes of hybrid-behavior vehicles; wherein the non-transitory computer readable medium stores instructions for creating different environment-based behavioral models for the different classes.
11. The non-transitory computer readable medium according to claim 10 wherein the analyzing comprises classifying the hybrid-behavior vehicles to the different classes based on visual attributes of the hybrid-behavior vehicles.
12. The non-transitory computer readable medium according to claim 10 wherein the analyzing comprises classifying the hybrid-behavior vehicles to the different classes based on movement parameters of the hybrid-behavior vehicles.
13. The non-transitory computer readable medium according to claim 10 wherein the analyzing comprises classifying the hybrid-behavior vehicles to the different classes based on risk levels associated with movements of the hybrid-behavior vehicles.
14. The non-transitory computer readable medium according to claim 10 wherein the analyzing comprises classifying the hybrid-behavior vehicles to the different classes based on at least two out of (a) visual attributes of the hybrid-behavior vehicles, (b) movement parameters of the hybrid-behavior vehicles, and (c) risk levels associated with movements of the hybrid-behavior vehicles.
15. A method for driving an autonomous vehicle, the method comprises:
- sensing an environment of the autonomous vehicle by one or more autonomous vehicle sensor;
- detecting, in the environment of the autonomous vehicle, environmental elements and a hybrid-behavior vehicle;
- predicting one or more predicted behaviors of the hybrid behavior vehicle based on the environmental elements and an environment-based behavioral model of the hybrid-behavior vehicle; wherein the environment-based behavioral model of the hybrid-behavior vehicle was generated by a computerized system configured to reduce consumption of memory and computational resources by preventing from building behavioral models of non-hybrid-behavior vehicles; wherein each one of the non-hybrid-behavior vehicles differs from the hybrid-behavior vehicle, wherein the hybrid-behavior vehicle is selected out of a group consisting of an unmotorized bicycle, a motorized bicycle, a motorized skateboard, a unicycle, a hoverboards and a kick scooter; and
- driving the autonomous vehicle based, at least in part, on the one or more predicted behaviors of the hybrid behavior vehicle.
16. The method according to claim 15 comprising predicting probabilities related to the occurrence of the one or more predicted behaviors of the hybrid behavior vehicle.
17. The method according to claim 15 wherein the driving comprises slowing down or stopping.
18. The method according to claim 15 wherein the predicting is responsive to a current behavior pattern of the hybrid-behavior vehicle.
19. A method for assisting a driver of a vehicle, the method comprises:
- sensing an environment of the vehicle by one or more vehicle sensor;
- detecting, in the environment of the vehicle, environmental elements and a hybrid-behavior vehicle;
- estimating an estimated future movement of the hybrid-behavior vehicle based on the environmental elements and an environment-based behavioral model of the hybrid-behavior vehicle; wherein the environment-based behavioral model of the hybrid-behavior vehicle was generated by a computerized system configured to reduce consumption of memory and computational resources by preventing from building behavioral models of non-hybrid-behavior vehicles; wherein each one of the non-hybrid-behavior vehicles differs from the hybrid-behavior vehicle, wherein the hybrid-behavior vehicle is selected out of a group consisting of an unmotorized bicycle, a motorized bicycle, a motorized skateboard, a unicycle, a hoverboards and a kick scooter; and
- performing at least one out of:
- providing to the driver an indication about the one or more predicted behaviors of the hybrid behavior vehicle to the driver; and
- calculating a suggested propagation path of the vehicle, based on the one or more predicted behaviors of the hybrid behavior vehicle, and providing to the driver an indication about the suggested propagation path of the vehicle.
20. The method according to claim 19 comprising predicting probabilities related to the occurrence of the one or more predicted behaviors of the hybrid behavior vehicle.
21. A non-transitory computer readable medium for driving an autonomous vehicle, the non-transitory computer readable medium that stores instructions for:
- sensing an environment of the autonomous vehicle by one or more autonomous vehicle sensor;
- detecting, in the environment of the autonomous vehicle, environmental elements and a hybrid-behavior vehicle; wherein each one of the non-hybrid-behavior vehicles differs from the hybrid-behavior vehicle, wherein the hybrid-behavior vehicle is selected out of a group consisting of an unmotorized bicycle, a motorized bicycle, a motorized skateboard, a unicycle, a hoverboards and a kick scooter;
- predicting one or more predicted behaviors of the hybrid behavior vehicle based on the environmental elements and an environment-based behavioral model of the hybrid-behavior vehicle; wherein the environment-based behavioral model of the hybrid-behavior vehicle was generated by a computerized system configured to reduce consumption of memory and computational resources by preventing from building behavioral models of non-hybrid-behavior vehicles; and
- driving the autonomous vehicle based, at least in part, on the one or more predicted behaviors of the hybrid behavior vehicle.
22. The non-transitory computer readable medium according to claim 21 that stores instructions for predicting probabilities related to the occurrence of the one or more predicted behaviors of the hybrid behavior vehicle.
23. The non-transitory computer readable medium according to claim 21 wherein the driving comprises slowing down or stopping.
24. The non-transitory computer readable medium according to claim 21, wherein the predicting is responsive to a current behavior pattern of the hybrid-behavior vehicle.
25. A non-transitory computer readable medium for assisting a driver of a vehicle, the non-transitory computer readable medium that stores instructions for:
- sensing an environment of the vehicle by one or more vehicle sensor; detecting, in the environment of the vehicle, environmental elements and a hybrid-behavior vehicle; estimating an estimated future movement of the hybrid-behavior vehicle based on the environmental elements and an environment-based behavioral model of the hybrid-behavior vehicle; wherein the environment-based behavioral model of the hybrid-behavior vehicle was generated by a computerized system configured to reduce consumption of memory and computational resources by preventing from building behavioral models of non-hybrid-behavior vehicles; wherein each one of the non-hybrid-behavior vehicles differs from the hybrid-behavior vehicle, wherein the hybrid-behavior vehicle is selected out of a group consisting of an unmotorized bicycle, a motorized bicycle, a motorized skateboard, a unicycle, a hoverboards and a kick scooter; and performing at least one out of: providing to the driver an indication about the one or more predicted behaviors of the hybrid behavior vehicle to the driver; and calculating a suggested propagation path of the vehicle, based on the one or more predicted behaviors of the hybrid behavior vehicle, and providing to the driver an indication about the suggested propagation path of the vehicle.
26. The non-transitory computer readable medium according to claim 25 that stores instructions for predicting probabilities related to the occurrence of the one or more predicted behaviors of the hybrid behavior vehicle.
4733353 | March 22, 1988 | Jaswa |
4932645 | June 12, 1990 | Schorey et al. |
4972363 | November 20, 1990 | Nguyen et al. |
5078501 | January 7, 1992 | Hekker et al. |
5214746 | May 25, 1993 | Fogel et al. |
5307451 | April 26, 1994 | Clark |
5412564 | May 2, 1995 | Ecer |
5436653 | July 25, 1995 | Ellis et al. |
5568181 | October 22, 1996 | Greenwood et al. |
5638425 | June 10, 1997 | Meador, III et al. |
5745678 | April 28, 1998 | Herzberg et al. |
5754938 | May 19, 1998 | Herz et al. |
5763069 | June 9, 1998 | Jordan |
5806061 | September 8, 1998 | Chaudhuri et al. |
5835087 | November 10, 1998 | Terz et al. |
5835901 | November 10, 1998 | Duvoisin et al. |
5852435 | December 22, 1998 | Vigneaux et al. |
5870754 | February 9, 1999 | Dimitrova et al. |
5873080 | February 16, 1999 | Coden et al. |
5887193 | March 23, 1999 | Takahashi et al. |
5926812 | July 20, 1999 | Hilsenrath et al. |
5978754 | November 2, 1999 | Kumano |
5991306 | November 23, 1999 | Burns et al. |
6052481 | April 18, 2000 | Grajski et al. |
6070167 | May 30, 2000 | Qian et al. |
6076088 | June 13, 2000 | Paik et al. |
6122628 | September 19, 2000 | Castelli et al. |
6128651 | October 3, 2000 | Cezar |
6137911 | October 24, 2000 | Zhilyaev |
6144767 | November 7, 2000 | Bottou et al. |
6147636 | November 14, 2000 | Gershenson |
6163510 | December 19, 2000 | Lee et al. |
6243375 | June 5, 2001 | Speicher |
6243713 | June 5, 2001 | Nelson et al. |
6275599 | August 14, 2001 | Adler et al. |
6314419 | November 6, 2001 | Faisal |
6329986 | December 11, 2001 | Cheng |
6381656 | April 30, 2002 | Shankman |
6411229 | June 25, 2002 | Kobayashi |
6422617 | July 23, 2002 | Fukumoto et al. |
6507672 | January 14, 2003 | Watkins et al. |
6523046 | February 18, 2003 | Liu et al. |
6524861 | February 25, 2003 | Anderson |
6546405 | April 8, 2003 | Gupta et al. |
6550018 | April 15, 2003 | Abonamah et al. |
6557042 | April 29, 2003 | He et al. |
6594699 | July 15, 2003 | Sahai et al. |
6601026 | July 29, 2003 | Appelt et al. |
6611628 | August 26, 2003 | Sekiguchi et al. |
6618711 | September 9, 2003 | Ananth |
6640015 | October 28, 2003 | Lafruit |
6643620 | November 4, 2003 | Contolini et al. |
6643643 | November 4, 2003 | Lee et al. |
6665657 | December 16, 2003 | Dibachi |
6681032 | January 20, 2004 | Bortolussi et al. |
6704725 | March 9, 2004 | Lee |
6732149 | May 4, 2004 | Kephart |
6742094 | May 25, 2004 | Igari |
6751363 | June 15, 2004 | Natsev et al. |
6751613 | June 15, 2004 | Lee et al. |
6754435 | June 22, 2004 | Kim |
6763069 | July 13, 2004 | Divakaran et al. |
6763519 | July 13, 2004 | McColl et al. |
6774917 | August 10, 2004 | Foote et al. |
6795818 | September 21, 2004 | Lee |
6804356 | October 12, 2004 | Krishnamachari |
6813395 | November 2, 2004 | Kinjo |
6819797 | November 16, 2004 | Smith et al. |
6877134 | April 5, 2005 | Fuller et al. |
6901207 | May 31, 2005 | Watkins |
6938025 | August 30, 2005 | Lulich et al. |
6985172 | January 10, 2006 | Rigney et al. |
7013051 | March 14, 2006 | Sekiguchi et al. |
7020654 | March 28, 2006 | Najmi |
7023979 | April 4, 2006 | Wu et al. |
7043473 | May 9, 2006 | Rassool et al. |
7158681 | January 2, 2007 | Persiantsev |
7215828 | May 8, 2007 | Luo |
7260564 | August 21, 2007 | Lynn et al. |
7289643 | October 30, 2007 | Brunk et al. |
7299261 | November 20, 2007 | Oliver et al. |
7302089 | November 27, 2007 | Smits |
7302117 | November 27, 2007 | Sekiguchi et al. |
7313805 | December 25, 2007 | Rosin et al. |
7340358 | March 4, 2008 | Yoneyama |
7346629 | March 18, 2008 | Kapur et al. |
7353224 | April 1, 2008 | Chen et al. |
7376672 | May 20, 2008 | Weare |
7383179 | June 3, 2008 | Alves et al. |
7433895 | October 7, 2008 | Li et al. |
7464086 | December 9, 2008 | Black et al. |
7529659 | May 5, 2009 | Wold |
7657100 | February 2, 2010 | Gokturk et al. |
7660468 | February 9, 2010 | Gokturk et al. |
7801893 | September 21, 2010 | Gulli |
7805446 | September 28, 2010 | Potok et al. |
7860895 | December 28, 2010 | Scofield et al. |
7872669 | January 18, 2011 | Darrell et al. |
7921288 | April 5, 2011 | Hildebrand |
7933407 | April 26, 2011 | Keidar et al. |
8023739 | September 20, 2011 | Tohimer et al. |
8266185 | September 11, 2012 | Raichelgauz et al. |
8275764 | September 25, 2012 | Jeon |
8285718 | October 9, 2012 | Ong et al. |
8312031 | November 13, 2012 | Raichelgauz et al. |
8315442 | November 20, 2012 | Gokturk et al. |
8345982 | January 1, 2013 | Gokturk et al. |
8386400 | February 26, 2013 | Raichelgauz et al. |
8396876 | March 12, 2013 | Kennedy et al. |
8418206 | April 9, 2013 | Bryant et al. |
RE44225 | May 21, 2013 | Aviv |
8442321 | May 14, 2013 | Chang et al. |
8457827 | June 4, 2013 | Ferguson et al. |
8495489 | July 23, 2013 | Everingham |
8527978 | September 3, 2013 | Sallam |
8634980 | January 21, 2014 | Urmson |
8635531 | January 21, 2014 | Graham et al. |
8655801 | February 18, 2014 | Raichelgauz et al. |
8655878 | February 18, 2014 | Kulkarni et al. |
8781152 | July 15, 2014 | Momeyer |
8782077 | July 15, 2014 | Rowley |
8799195 | August 5, 2014 | Raichelgauz et al. |
8799196 | August 5, 2014 | Raichelquaz et al. |
8818916 | August 26, 2014 | Raichelgauz et al. |
8868861 | October 21, 2014 | Shimizu et al. |
8886648 | November 11, 2014 | Procopio et al. |
8954887 | February 10, 2015 | Tseng et al. |
8990199 | March 24, 2015 | Ramesh et al. |
9009086 | April 14, 2015 | Raichelgauz et al. |
9104747 | August 11, 2015 | Raichelgauz et al. |
9165406 | October 20, 2015 | Gray et al. |
9298763 | March 29, 2016 | Zack |
9311308 | April 12, 2016 | Sankarasubramaniam et al. |
9323754 | April 26, 2016 | Ramanathan et al. |
9440647 | September 13, 2016 | Sucan |
9466068 | October 11, 2016 | Raichelgauz et al. |
9646006 | May 9, 2017 | Raichelgauz et al. |
9679062 | June 13, 2017 | Schillings et al. |
9734533 | August 15, 2017 | Givot |
9807442 | October 31, 2017 | Bhatia et al. |
9875445 | January 23, 2018 | Amer et al. |
9984369 | May 29, 2018 | Li et al. |
10133947 | November 20, 2018 | Yang |
10347122 | July 9, 2019 | Takenaka |
10491885 | November 26, 2019 | Hicks |
20010019633 | September 6, 2001 | Tenze et al. |
20010034219 | October 25, 2001 | Hewitt et al. |
20010038876 | November 8, 2001 | Anderson |
20020004743 | January 10, 2002 | Kutaragi et al. |
20020010682 | January 24, 2002 | Johnson |
20020010715 | January 24, 2002 | Chinn et al. |
20020019881 | February 14, 2002 | Bokhari et al. |
20020032677 | March 14, 2002 | Morgenthaler et al. |
20020038299 | March 28, 2002 | Zernik et al. |
20020042914 | April 11, 2002 | Walker et al. |
20020072935 | June 13, 2002 | Rowse et al. |
20020087530 | July 4, 2002 | Smith et al. |
20020087828 | July 4, 2002 | Arimilli et al. |
20020091947 | July 11, 2002 | Nakamura |
20020107827 | August 8, 2002 | Benitez-Jimenez et al. |
20020113812 | August 22, 2002 | Walker et al. |
20020126002 | September 12, 2002 | Patchell |
20020126872 | September 12, 2002 | Brunk et al. |
20020129140 | September 12, 2002 | Peled et al. |
20020147637 | October 10, 2002 | Kraft et al. |
20020157116 | October 24, 2002 | Jasinschi |
20020163532 | November 7, 2002 | Thomas et al. |
20020174095 | November 21, 2002 | Lulich et al. |
20020184505 | December 5, 2002 | Mihcak et al. |
20030004966 | January 2, 2003 | Bolle et al. |
20030005432 | January 2, 2003 | Ellis et al. |
20030037010 | February 20, 2003 | Schmelzer |
20030041047 | February 27, 2003 | Chang et al. |
20030089216 | May 15, 2003 | Birmingham et al. |
20030093790 | May 15, 2003 | Logan et al. |
20030101150 | May 29, 2003 | Agnihotri et al. |
20030105739 | June 5, 2003 | Essafi et al. |
20030110236 | June 12, 2003 | Yang et al. |
20030115191 | June 19, 2003 | Copperman et al. |
20030126147 | July 3, 2003 | Bsafi et al. |
20030140257 | July 24, 2003 | Peterka et al. |
20030165269 | September 4, 2003 | Fedorovskaya et al. |
20030174859 | September 18, 2003 | Kim |
20030184598 | October 2, 2003 | Graham |
20030200217 | October 23, 2003 | Ackerman |
20030217335 | November 20, 2003 | Chung et al. |
20030229531 | December 11, 2003 | Heckerman et al. |
20040059736 | March 25, 2004 | Willse |
20040091111 | May 13, 2004 | Levy |
20040095376 | May 20, 2004 | Graham et al. |
20040098671 | May 20, 2004 | Graham et al. |
20040111432 | June 10, 2004 | Adams et al. |
20040117638 | June 17, 2004 | Monroe |
20040128511 | July 1, 2004 | Sun et al. |
20040153426 | August 5, 2004 | Nugent |
20040162820 | August 19, 2004 | James et al. |
20040230572 | November 18, 2004 | Omoigui |
20040267774 | December 30, 2004 | Lin et al. |
20050021394 | January 27, 2005 | Miedema et al. |
20050080788 | April 14, 2005 | Murata |
20050114198 | May 26, 2005 | Koningstein et al. |
20050131884 | June 16, 2005 | Gross et al. |
20050163375 | July 28, 2005 | Grady |
20050172130 | August 4, 2005 | Roberts |
20050177372 | August 11, 2005 | Wang et al. |
20050193015 | September 1, 2005 | Logston |
20050226511 | October 13, 2005 | Short |
20050238198 | October 27, 2005 | Brown et al. |
20050238238 | October 27, 2005 | Xu et al. |
20050249398 | November 10, 2005 | Khamene et al. |
20050256820 | November 17, 2005 | Dugan et al. |
20050262428 | November 24, 2005 | Little et al. |
20050281439 | December 22, 2005 | Lange |
20050289163 | December 29, 2005 | Gordon et al. |
20050289590 | December 29, 2005 | Cheok et al. |
20060004745 | January 5, 2006 | Kuhn et al. |
20060015580 | January 19, 2006 | Gabriel et al. |
20060020958 | January 26, 2006 | Allamanche et al. |
20060033163 | February 16, 2006 | Chen |
20060050993 | March 9, 2006 | Stentiford |
20060069668 | March 30, 2006 | Braddy et al. |
20060080311 | April 13, 2006 | Potok et al. |
20060100987 | May 11, 2006 | Leurs |
20060112035 | May 25, 2006 | Cecchi et al. |
20060120626 | June 8, 2006 | Perlmutter |
20060129822 | June 15, 2006 | Snijder et al. |
20060217818 | September 28, 2006 | Fujiwara |
20060217828 | September 28, 2006 | Hicken |
20060218191 | September 28, 2006 | Gopalakrishnan |
20060224529 | October 5, 2006 | Kermani |
20060236343 | October 19, 2006 | Chang |
20060242130 | October 26, 2006 | Sadri et al. |
20060248558 | November 2, 2006 | Barton et al. |
20060251338 | November 9, 2006 | Gokturk et al. |
20060251339 | November 9, 2006 | Gokturk |
20060253423 | November 9, 2006 | McLane et al. |
20060288002 | December 21, 2006 | Epstein et al. |
20070022374 | January 25, 2007 | Huang et al. |
20070033170 | February 8, 2007 | Sull et al. |
20070038614 | February 15, 2007 | Guha |
20070042757 | February 22, 2007 | Jung et al. |
20070061302 | March 15, 2007 | Ramer et al. |
20070067304 | March 22, 2007 | Ives |
20070074147 | March 29, 2007 | Wold |
20070083611 | April 12, 2007 | Farago et al. |
20070091106 | April 26, 2007 | Moroney |
20070130159 | June 7, 2007 | Gulli et al. |
20070136782 | June 14, 2007 | Ramaswamy et al. |
20070156720 | July 5, 2007 | Maren |
20070196013 | August 23, 2007 | Li |
20070244902 | October 18, 2007 | Seide et al. |
20070253594 | November 1, 2007 | Lu et al. |
20070298152 | December 27, 2007 | Baets |
20080049789 | February 28, 2008 | Vedantham et al. |
20080072256 | March 20, 2008 | Boicey et al. |
20080079729 | April 3, 2008 | Brailovsky |
20080109433 | May 8, 2008 | Rose |
20080152231 | June 26, 2008 | Gokturk et al. |
20080159622 | July 3, 2008 | Agnihotri et al. |
20080165861 | July 10, 2008 | Wen et al. |
20080166020 | July 10, 2008 | Kosaka |
20080201299 | August 21, 2008 | Lehikoinen et al. |
20080201314 | August 21, 2008 | Smith et al. |
20080201361 | August 21, 2008 | Castro et al. |
20080228995 | September 18, 2008 | Tan et al. |
20080237359 | October 2, 2008 | Silverbrook et al. |
20080247543 | October 9, 2008 | Mick et al. |
20080253737 | October 16, 2008 | Kimura et al. |
20080263579 | October 23, 2008 | Mears et al. |
20080270373 | October 30, 2008 | Oostveen et al. |
20080270569 | October 30, 2008 | McBride |
20080294278 | November 27, 2008 | Borgeson et al. |
20080307454 | December 11, 2008 | Ahanger et al. |
20080313140 | December 18, 2008 | Pereira et al. |
20090022472 | January 22, 2009 | Bronstein |
20090024641 | January 22, 2009 | Quigley et al. |
20090034791 | February 5, 2009 | Doretto |
20090037088 | February 5, 2009 | Taguchi |
20090043637 | February 12, 2009 | Eder |
20090043818 | February 12, 2009 | Raichelgauz |
20090080759 | March 26, 2009 | Bhaskar |
20090096634 | April 16, 2009 | Emam et al. |
20090125544 | May 14, 2009 | Brindley |
20090157575 | June 18, 2009 | Schobben et al. |
20090165031 | June 25, 2009 | Li et al. |
20090172030 | July 2, 2009 | Schiff et al. |
20090208106 | August 20, 2009 | Dunlop et al. |
20090208118 | August 20, 2009 | Csurka |
20090216761 | August 27, 2009 | Raichelgauz et al. |
20090220138 | September 3, 2009 | Zhang et al. |
20090245573 | October 1, 2009 | Saptharishi et al. |
20090254572 | October 8, 2009 | Redlich et al. |
20090278934 | November 12, 2009 | Ecker |
20090282218 | November 12, 2009 | Raichelgauz et al. |
20090297048 | December 3, 2009 | Slotine et al. |
20100042646 | February 18, 2010 | Raichelgauz et al. |
20100082684 | April 1, 2010 | Churchill et al. |
20100104184 | April 29, 2010 | Bronstein et al. |
20100111408 | May 6, 2010 | Matsuhira |
20100125569 | May 20, 2010 | Nair et al. |
20100162405 | June 24, 2010 | Cook et al. |
20100191391 | July 29, 2010 | Zeng |
20100198626 | August 5, 2010 | Cho et al. |
20100212015 | August 19, 2010 | Jin et al. |
20100284604 | November 11, 2010 | Chrysanthakopoulos |
20100293057 | November 18, 2010 | Haveliwala et al. |
20100306193 | December 2, 2010 | Pereira |
20100312736 | December 9, 2010 | Kello |
20100318493 | December 16, 2010 | Wessling |
20100325138 | December 23, 2010 | Lee et al. |
20100325581 | December 23, 2010 | Finkelstein et al. |
20110029620 | February 3, 2011 | Bonforte |
20110035373 | February 10, 2011 | Berg et al. |
20110038545 | February 17, 2011 | Bober |
20110055585 | March 3, 2011 | Lee |
20110164180 | July 7, 2011 | Lee |
20110164810 | July 7, 2011 | Zang et al. |
20110216209 | September 8, 2011 | Fredlund et al. |
20110218946 | September 8, 2011 | Stern et al. |
20110246566 | October 6, 2011 | Kashef |
20110276680 | November 10, 2011 | Rimon |
20110296315 | December 1, 2011 | Lin et al. |
20120131454 | May 24, 2012 | Shah |
20120133497 | May 31, 2012 | Sasaki |
20120136853 | May 31, 2012 | Kennedy et al. |
20120167133 | June 28, 2012 | Carroll et al. |
20120179642 | July 12, 2012 | Sweeney et al. |
20120179751 | July 12, 2012 | Ahn |
20120185445 | July 19, 2012 | Borden et al. |
20120207346 | August 16, 2012 | Kohli et al. |
20120221470 | August 30, 2012 | Lyon |
20120227074 | September 6, 2012 | Hill et al. |
20120239690 | September 20, 2012 | Asikainen et al. |
20120239694 | September 20, 2012 | Avner et al. |
20120265735 | October 18, 2012 | McMillan et al. |
20120294514 | November 22, 2012 | Saunders et al. |
20120299961 | November 29, 2012 | Ramkumar et al. |
20120301105 | November 29, 2012 | Rehg et al. |
20120331011 | December 27, 2012 | Raichelgauz et al. |
20130043990 | February 21, 2013 | Al-Jafar |
20130066856 | March 14, 2013 | Ong et al. |
20130067364 | March 14, 2013 | Berntson et al. |
20130086499 | April 4, 2013 | Dyor et al. |
20130089248 | April 11, 2013 | Remiszewski et al. |
20130103814 | April 25, 2013 | Carrasco |
20130151522 | June 13, 2013 | Aggarwal et al. |
20130159298 | June 20, 2013 | Mason et al. |
20130212493 | August 15, 2013 | Krishnamurthy |
20130226820 | August 29, 2013 | Sedota, Jr. |
20130226930 | August 29, 2013 | Arngren et al. |
20130227023 | August 29, 2013 | Raichelgauz et al. |
20130283401 | October 24, 2013 | Pabla et al. |
20130346412 | December 26, 2013 | Raichelgauz et al. |
20140019264 | January 16, 2014 | Wachman et al. |
20140025692 | January 23, 2014 | Pappas |
20140032015 | January 30, 2014 | Chun |
20140059443 | February 27, 2014 | Tabe |
20140095425 | April 3, 2014 | Sipple |
20140111647 | April 24, 2014 | Atsmon |
20140125703 | May 8, 2014 | Roveta et al. |
20140147829 | May 29, 2014 | Jerauld |
20140149918 | May 29, 2014 | Asokan et al. |
20140152698 | June 5, 2014 | Kim et al. |
20140156691 | June 5, 2014 | Conwell |
20140169681 | June 19, 2014 | Drake |
20140176604 | June 26, 2014 | Venkitaraman et al. |
20140193077 | July 10, 2014 | Shiiyama et al. |
20140198986 | July 17, 2014 | Marchesotti |
20140201330 | July 17, 2014 | Lopez et al. |
20140250032 | September 4, 2014 | Huang et al. |
20140282655 | September 18, 2014 | Roberts |
20140300722 | October 9, 2014 | Garcia |
20140330830 | November 6, 2014 | Raichelgauz et al. |
20140341476 | November 20, 2014 | Kulick et al. |
20140363044 | December 11, 2014 | Williams et al. |
20140379477 | December 25, 2014 | Sheinfeld |
20150033150 | January 29, 2015 | Lee |
20150052089 | February 19, 2015 | Kozloski et al. |
20150100562 | April 9, 2015 | Kohlmeier et al. |
20150117784 | April 30, 2015 | Lin et al. |
20150120627 | April 30, 2015 | Hunzinger et al. |
20150127516 | May 7, 2015 | Studnitzer et al. |
20150134688 | May 14, 2015 | Jing |
20150248586 | September 3, 2015 | Gaidon et al. |
20150254344 | September 10, 2015 | Kulkarni et al. |
20150266455 | September 24, 2015 | Wilson |
20150286742 | October 8, 2015 | Zhang et al. |
20150286872 | October 8, 2015 | Medioni et al. |
20150324356 | November 12, 2015 | Gutierrez et al. |
20150332588 | November 19, 2015 | Bulan et al. |
20150363644 | December 17, 2015 | Wnuk |
20160007083 | January 7, 2016 | Gurha |
20160026707 | January 28, 2016 | Ong et al. |
20160132194 | May 12, 2016 | Grue et al. |
20160210525 | July 21, 2016 | Yang |
20160221592 | August 4, 2016 | Puttagunta et al. |
20160275766 | September 22, 2016 | Venetianer et al. |
20160306798 | October 20, 2016 | Guo et al. |
20160342683 | November 24, 2016 | Kwon |
20160357188 | December 8, 2016 | Ansari |
20170017638 | January 19, 2017 | Satyavarta et al. |
20170032257 | February 2, 2017 | Sharifi |
20170041254 | February 9, 2017 | Agara Venkatesha Rao |
20170109602 | April 20, 2017 | Kim |
20170154241 | June 1, 2017 | Shambik et al. |
20170255620 | September 7, 2017 | Raichelgauz |
20170262437 | September 14, 2017 | Raichelgauz |
20170323568 | November 9, 2017 | Inoue |
20180081368 | March 22, 2018 | Watanabe |
20180101177 | April 12, 2018 | Cohen |
20180108258 | April 19, 2018 | Dilger |
20180157903 | June 7, 2018 | Tu et al. |
20180157916 | June 7, 2018 | Doumbouya |
20180158323 | June 7, 2018 | Takenaka |
20180189613 | July 5, 2018 | Wolf et al. |
20180204111 | July 19, 2018 | Zadeh |
20180373929 | December 27, 2018 | Ye |
20190005726 | January 3, 2019 | Nakano |
20190039627 | February 7, 2019 | Yamamoto |
20190043274 | February 7, 2019 | Hayakawa |
20190045244 | February 7, 2019 | Balakrishnan |
20190056718 | February 21, 2019 | Satou |
20190065951 | February 28, 2019 | Luo |
20190071093 | March 7, 2019 | Ma |
20190072965 | March 7, 2019 | Zhang |
20190096135 | March 28, 2019 | Mutto et al. |
20190171912 | June 6, 2019 | Vallespi-Gonzalez et al. |
20190188501 | June 20, 2019 | Ryu |
20190220011 | July 18, 2019 | Della Penna |
20190279046 | September 12, 2019 | Han et al. |
20190304102 | October 3, 2019 | Chen et al. |
20190311226 | October 10, 2019 | Xiao |
20190317513 | October 17, 2019 | Zhang |
20190364492 | November 28, 2019 | Azizi |
20190384303 | December 19, 2019 | Muller |
20190384312 | December 19, 2019 | Herbach |
20190385460 | December 19, 2019 | Magzimof |
20190389459 | December 26, 2019 | Berntorp |
20200004248 | January 2, 2020 | Healey |
20200004251 | January 2, 2020 | Zhu |
20200004265 | January 2, 2020 | Zhu |
20200005631 | January 2, 2020 | Visintainer |
20200018606 | January 16, 2020 | Wolcott |
20200018618 | January 16, 2020 | Ozog |
20200020212 | January 16, 2020 | Song |
20200050973 | February 13, 2020 | Stenneth |
20200073977 | March 5, 2020 | Montemerlo |
20200074326 | March 5, 2020 | Balakrishnan |
20200090484 | March 19, 2020 | Chen |
20200097756 | March 26, 2020 | Hashimoto |
20200133307 | April 30, 2020 | Kelkar |
20200043326 | February 6, 2020 | Tao |
1085464 | January 2007 | EP |
0231764 | April 2002 | WO |
2003067467 | August 2003 | WO |
2005027457 | March 2005 | WO |
2007049282 | May 2007 | WO |
2014076002 | May 2014 | WO |
2014137337 | September 2014 | WO |
2016040376 | March 2016 | WO |
2016070193 | May 2016 | WO |
- Zhou et al, “Ensembling neural networks: Many could be better than all”, National Laboratory for Novel Software Technology, Nanjing University, Hankou Road 22, Nanjing 210093, PR China Received Nov. 16, 2001, Available online Mar. 12, 2002, pp. 239-263.
- Zhou et al, “Medical Diagnosis With C4.5 Rule Preceded by Artificial Neural Network Ensemble”, IEEE Transactions on Information Technology in Biomedicine, vol. 7, Issue: 1, Mar. 2003, pp. 37-42.
- Zhu et al., “Technology-Assisted Dietary Assesment”, Proc SPIE. Mar. 20, 2008, pp. 1-15.
- Zou et al., “A Content-Based Image Authentication System with Lossless Data Hiding”, ICME 2003, pp. 213-216.
- “Computer Vision Demonstration Website”, Electronics and Computer Science, University of Southampton, 2005, USA.
- Big Bang Theory Series 04 Episode 12, aired Jan. 6, 2011; [retrieved from Internet: ].
- Boari et al, “Adaptive Routing for Dynamic Applications in Massively Parallel Architectures”, 1995 IEEE, Spring 1995, pp. 1-14.
- Burgsteiner et al., “Movement Prediction from Real-World Images Using a Liquid State machine”, Innovations in Applied Artificial Intelligence Lecture Notes in Computer Science, Lecture Notes in Artificial Intelligence, LNCS, Springer-Verlag, BE, vol. 3533, Jun. 2005, pp. 121-130.
- Cernansky et al, “Feed-forward Echo State Networks”, Proceedings of International Joint Conference on Neural Networks, Montreal, Canada, Jul. 31-Aug. 4, 2005, pp. 1-4.
- Chinchor, Nancy A. et al.; Multimedia Analysis + Visual Analytics = Multimedia Analytics; IEEE Computer Society; 2010; pp. 52-60. (Year: 2010).
- Fathy et al, “A Parallel Design and Implementation For Backpropagation Neural Network Using MIMD Architecture”, 8th Mediterranean Electrotechnical Conference, 19'96. MELECON '96, Date of Conference: May 13-16, 1996, vol. 3 pp. 1472-1475, vol. 3.
- Freisleben et al, “Recognition of Fractal Images Using a Neural Network”, Lecture Notes in Computer Science, 1993, vol. 6861, 1993, pp. 631-637.
- Garcia, “Solving the Weighted Region Least Cost Path Problem Using Transputers”, Naval Postgraduate School, Monterey, California, Dec. 1989.
- Guo et al, AdOn: An Intelligent Overlay Video Advertising System (Year: 2009).
- Hogue, “Tree Pattern Inference and Matching for Wrapper Induction on the World Wide Web”, Master's Thesis, Massachusetts Institute of Technology, Jun. 2004, pp. 1-106.
- Howlett et al, “A Multi-Computer Neural Network Architecture in a Virtual Sensor System Application”, International journal of knowledge-based intelligent engineering systems, 4 (2). pp. 86-93, 133N 1327-2314.
- Hua et al., “Robust Video Signature Based on Ordinal Measure”, Image Processing, 2004, 2004 International Conference on Image Processing (ICIP), vol. 1, IEEE, pp. 685-688, 2004.
- International Search Report and Written Opinion for PCT/US2016/050471, ISA/RU, Moscow, RU, dated May 4, 2017.
- International Search Report and Written Opinion for PCT/US2016/054634, ISA/RU, Moscow, RU, dated Mar. 16, 2017.
- International Search Report and Written Opinion for PCT/US2017/015831, ISA/RU, Moscow, RU, dated Apr. 20, 2017.
- Johnson et al., “Pulse-Coupled Neural Nets: Translation, Rotation, Scale, Distortion, and Intensity Signal Invariance for Images”, Applied Optics, vol. 33, No. 26, 1994, pp. 6239-6253.
- Lau et al., “Semantic Web Service Adaptation Model for a Pervasive Learning Scenario”, 2008 IEEE Conference on Innovative Technologies in Intelligent Systems and Industrial Applications, 2008, pp. 98-103.
- Li et al (“Matching Commercial Clips from TV Streams Using a Unique, Robust and Compact Signature” 2005) (Year: 2005).
- Lin et al., “Generating robust digital signature for image/video authentication”, Multimedia and Security Workshop at ACM Multimedia '98, Bristol, U.K., Sep. 1998, pp. 245-251.
- Lu et al, “Structural Digital Signature for Image Authentication: An Incidental Distortion Resistant Scheme”, IEEE Transactions on Multimedia, vol. 5, No. 2, Jun. 2003, pp. 161-173.
- Lyon, “Computational Models of Neural Auditory Processing”, IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP '84, Date of Conference: Mar. 1984, vol. 9, pp. 41-44.
- Ma Et El “Semantics modeling based image retrieval system using neural networks”, 2005.
- Marian Stewart B et al., “Independent component representations for face recognition”, Proceedings of the SPIE Symposium on Electronic Imaging: Science and Technology; Conference on Human Vision and Electronic Imaging III, San Jose, California, Jan. 1998, pp. 1-12.
- May et al, “The Transputer”, Springer-Verlag Berlin Heidelberg 1989, vol. 41.
- McNamara et al., “Diversity Decay in opportunistic Content Sharing Systems”, 2011 IEEE International Symposium on a World of Wireless, Mobile and Multimedia Networks, pp. 1-3.
- Morad et al., “Performance, Power Efficiency and Scalability of Asymmetric Cluster Chip Multiprocessors”, Computer Architecture Letters, vol. 4, Jul. 4, 2005, pp. 1-4, XP002466254.
- Nagy et al, “A Transputer, Based, Flexible, Real-Time Control System for Robotic Manipulators”, UKACC International Conference on CONTROL '96, Sep. 2-5, 1996, Conference Publication No. 427, IEE 1996.
- Natschlager et al., “The “Liquid Computer”: A novel strategy for real-time computing on time series”, Special Issue on Foundations of Information Processing of telematik, vol. 8, No. 1, 2002, pp. 39-43, XP002466253.
- Odinaev et al, “Cliques in Neural Ensembles as Perception Carriers”, Technion—Institute of Technology, 2006 International Joint Conference on neural Networks, Canada, 2006, pp. 285-292.
- Ortiz-Boyer et al, “CIXL2: A Crossover Operator for Evolutionary Algorithms Based on Population Features”, Journal of Artificial Intelligence Research 24 (2005) Submitted 11 /04; published Jul. 2005, pp. 1-48.
- Pandya etal. A Survey on QR Codes: in context of Research and Application. International Journal of Emerging Technology and U Advanced Engineering. ISSN 2250-2459, ISO 9001:2008 Certified Journal, vol. 4, Issue 3, Mar. 2014 (Year: 2014).
- Queluz, “Content-Based Integrity Protection of Digital Images”, SPIE Conf. on Security and Watermarking of Multimedia Contents, San Jose, Jan. 1999, pp. 85-93.
- Rui, Yong et al. “Relevance feedback: a power tool for interactive content-based image retrieval.” IEEE Transactions on circuits and systems for video technology 8.5 (1998): 644-655.
- Santos et al., “SCORM-MPEG: an Ontology of Interoperable Metadata for multimediaand E-Learning”, 23rd International Conference on Software, Telecommunications and Computer Networks (SoftCom), 2015, pp. 224-228.
- Scheper et al, “Nonlinear dynamics in neural computation”, ESANN'2006 proceedings—European Symposium on Artificial Neural Networks, Bruges (Belgium), Apr. 26-28, 2006, d-side publication, ISBN 2-930307-06-4, pp. 1-12.
- Schneider et al., “A Robust Content based Digital Signature for Image Authentication”, Proc. ICIP 1996, Lausane, Switzerland, Oct. 1996, pp. 227-230.
- Srihari et al., “Intelligent Indexing and Semantic Retrieval of Multimodal Documents”, Kluwer Academic Publishers, May 2000, vol. 2, Issue 2-3, pp. 245-275.
- Srihari, Rohini K. “Automatic indexing and content-based retrieval of captioned images” Computer 0 (1995): 49-56.
- Stolberg et al (“Hibrid-Soc: A Multi-Core Soc Architecture for Multimedia Signal Processing” (2003).
- Stolberg et al, “Hibrid-Soc: A Mul Ti-Core Soc Architecture for Mul Timedia Signal Processing”, 2003 IEEE, pp. 189-194.
- Theodoropoulos et al, “Simulating Asynchronous Architectures on Transputer Networks”, Proceedings of the Fourth Euromicro Workshop On Parallel and Distributed Processing, 1996. PDP '96, pp. 274-281.
- Vallet et al (“Personalized Content Retrieval in Context Using Ontological Knowledge” Mar. 2007) (Year: 2007).
- Verstraeten et al, “Isolated word recognition with the Liquid State Machine: a case study”, Department of Electronics and Information Systems, Ghent University, Sint-Pietersnieuwstraat 41, 9000 Gent, Belgium, Available onlline Jul. 14, 2005, pp. 521-528.
- Wang et al., “Classifying Objectionable Websites Based onlmage Content”, Stanford University, pp. 1-12.
- Ware et al, “Locating and Identifying Components in a Robot's Workspace using a Hybrid Computer Architecture” Proceedings of the 1995 IEEE International Symposium on Intelligent Control, Aug. 27-29, 1995, pp. 139-144.
- Whitby-Strevens, “The transputer”, 1985 IEEE, pp. 292-300.
- Wilk et al., “The Potential of Social-Aware Multimedia Prefetching on Mobile Devices”, International Conference and Workshops on networked Systems (NetSys), 2015, pp. 1-5.
- Yanagawa et al., “Columbia University's Baseline Detectors for 374 LSCOM Semantic Visual Concepts”, Columbia University ADVENT Technical Report # 222-2006-8, Mar. 20, 2007, pp. 1-17.
- Yanagawa et al., “Columbia University's Baseline Detectors for 374 LSCOM Semantic Visual Concepts”, Columbia University ADVENT Technical Report #222, 2007, pp. 2006-2008.
- Jasinschi et al., “A Porbabilistic Layered Framework for Integrating Multimedia Content and Context Information”, IEEE, p. 2057-2060. (Year: 2002).
- Jones et al., “Contextual Dynamics of Group-Based Sharing Decisions”, 2011, University of Bath, p. 1777-1786. (Year: 2011).
- Iwamoto, “Image Signature Robust to Caption Superimposition for Video Sequence Identification”, IEEE, pp. 3185-3188 (Year: 2006).
- Cooperative Multi-Scale Convolutional Neural, Networks for Person Detection, Markus Eisenbach, Daniel Seichter, Tim Wengefeld, and Horst-Michael Gross Ilmenau Universit of Technology, Neuroinformatics and Cognitive Robotics Lab (Year; 2016).
- Chen, Yixin, James Ze Wang, and Robert Krovetz. “CLUE: cluster-based retrieval of images by unsupervised learning.” IEEE transactions on Image Processing 14.8 (2005); 1187-1201. (Year: 2005).
- Wusk et al (Non-Invasive detection of Respiration and Heart Rate with a Vehicle Seat Sensor; www.mdpi.com/journal/sensors; Published: May 8, 2018). (Year: 2018).
- Chen, Tiffany Yu-Han, et al. “Glimpse: Continuous, real-time object recognition on mobile devices.” Proceedings of the 13th ACM Confrecene on Embedded Networked Sensor Systems. 2015. (Year: 2015).
Type: Grant
Filed: Mar 27, 2020
Date of Patent: Aug 6, 2024
Patent Publication Number: 20200310458
Assignee: AUTOBRAINS TECHNOLOGIES LTD (Tel-Aviv-Yafo)
Inventor: Igal Raichelgauz (Tel Aviv)
Primary Examiner: Abby Y Lin
Assistant Examiner: Renee LaRose
Application Number: 16/831,841
International Classification: G01C 21/36 (20060101); G05D 1/00 (20060101); G05D 1/02 (20200101); G06V 10/764 (20220101); G06V 20/40 (20220101); G06V 20/52 (20220101); G06V 20/58 (20220101); G06V 40/20 (20220101);